Contents lists available at ScienceDirect

Nano Energy

journal homepage: www.elsevier.com/locate/nanoen

Full paper

Self-powered nanodevices for fast UV detection and energy harvesting usingcore-shell nanowire geometry

Chun-Ho Lin1, Hui-Chun Fu1, Der-Hsien Lien1, Chia-Yang Hsu, Jr-Hau He⁎

Computer, Electrical, and Mathematical Sciences and Engineering (CEMSE) Division, King Abdullah University of Science & Technology (KAUST), Thuwal 23955-6900,Saudi Arabia

A R T I C L E I N F O

Keywords:Self-poweredCore-shellZnOPhotodetectorPhotovoltaic

A B S T R A C T

Developing multiple functionalities in nanodevices is a desirable objective in current nanotechnology, where theself-powered capability is one of the most demanded functions—without the need of external batteries, nano-devices can be operated independently and maintenance-free for a long period of time. In this study, we fab-ricated the core-shell nanostructured device employing multi-walled carbon nanotubes (CNTs) as inner cores andZnO as outer shells, and demonstrate their multiple functionalities of fast UV detection and self-powered cap-ability. The device features high photoconductive gain of 800 and short response time of 0.48ms in air and29ms in vacuum. For energy harvesting, the device shows photovoltaic (PV) capability to drive a conventionalliquid crystal display (LCD) with a Voc of 0.85 V and Isc of 5 nA. Those features demonstrate that our CNT-ZnOcore-shell nanostructured device, performing both high photoresponsivity and PV capability, can be operated indifferent functionalities with or without external bias.

1. Introduction

The developing of hierarchical structures by combining materials ofdifferent scales, classes, and electrical/optical properties has paved theway for tunable performance of nano-optoelectronics [1–4]. In the viewof optics, performance of nano-optoelectronics can be enhanced byemploying hierarchical structure which supports a sophisticated man-agement of light. For example, a combination of microgroove and NWstructure can effectively reduce the reflection loss of monocrystalline Sisolar cells [5], and three-dimensional dual-diameter nanopillars haveshown improved light absorption [6,7]. From the electrical viewpoint,carrier transports as well as photon-electron couplings within materialscan be facilitated through the heterojunctions introduced by hier-archical structures. For example, Hsu et al. reported the integration ofcarbon nanotubes and TiO2 shells to form radial Schottky barriers in acore-shell fashion to enhance the sensitivity and speed of the NWphotodetectors (PDs) [8]. Lieber et al. and Yang et al. created the core-shell nanowires as a platform to explore PV applications in nanoscale[9,10]. Implementation of hierarchical structures also helps to elim-inate the detrimental effects caused by the scaling down of the device,such as device integration difficulties and the scarification of deviceperformance. For example, in this study, through appropriate design ofhierarchical structures, the PD device can be powered by itself without

integrating additional power supply. The hierarchical structures alsoprovide a way to reinforce the performance of nanoelectronics. TakingZnO as an example, although their nano-enabled PDs exhibit superioroptoelectronic properties (e.g., the photoconductive gain can be up to~ 108) [11], the temporal resolution (on timescales of a few seconds tominutes) is still unsatisfactory due to an inevitable compromise be-tween sensitivity and speed [12,13]. Those faulty features limit theopportunities of nanodevices for real-time sensing and preclude theirapplications for various tasks. To improve the response speed, decora-tions of metallic or dielectric nanoparticles on metal-oxide NWs to formSchottky or pn junctions have shown remarkable effects [14]. Hier-archical structure, such as size selection [15], network schemes [16],biaxial [17], and core-shell structures [18], provide promising routesfor enhancing the sensitivity as well as the response/recovery speed.However, despite all of these achievements, beating the sensitivity-speed tradeoff remains a challenge.

Harvesting renewable energies from environments to realize a self-powered nanosystem has drawn significant attention in recent yearsdue to the energy and global warming issues. Such nanosystems arecomposed of energy-harvesting devices, such as solar cell, thermo-electric generator and nanogenerators to provide power, coupling withfunctional devices, like LEDs, PDs and various types of sensors to sup-port functionality [19–21]. However, although the integration between

https://doi.org/10.1016/j.nanoen.2018.06.065Received 16 May 2018; Received in revised form 9 June 2018; Accepted 19 June 2018

⁎ Corresponding author.

1 Equal contribution.E-mail address: jrhau.he@kaust.edu.sa (J.-H. He).

Nano Energy 51 (2018) 294–299

Available online 19 June 20182211-2855/ © 2018 Published by Elsevier Ltd.

T

devices enables the nanosystem to operate independently, powermatching, efficiency and scaling are still challenging. To get around themismatch, a current trend is to adapt active nanodevices that not onlydetect the signals but also capable of being powered by these detectedsignals. For example, Chen et al. have fabricated an active vibrationsensor able to be powered by the detected vibrations [22]. Wu et al.demonstrated an piezotronic strain sensor able to be powered by theexerted forces [23]. To develop the active PDs, it is important to exploitthe PV effect so as to realize a self-powered active sensor because it doesnot need an external power source to drive.

In this study, we design the core-shell nanostructures employingCNTs as inner cores and ZnO as outer shells (Fig. 1a) to realize the fastUV detection and self-powered capability. The employment of the coreCNT provides two key advantages: (i) providing additional conductionpath and reducing the need of ZnO thickness, and (ii) creating addi-tional junction with Schottky contacts. We show that those advantagescan greatly promote the response speed, enabling fast detection even invacuum conditions, which is not achievable for conventional ZnO-based PDs without hierarchical structures. Due to the existence of radialdepletion region extending along the NW, the device also exhibits PVcharacteristic and thus the UV energy can be converted to electricity.Owing to its small dimension and the use of asymmetric metal contacts(one side Ohmic and the other Schottky), the PD also features a verylow dark current, can work at zero bias, and is expected to have veryhigh operation speed. Furthermore, we show an example of how toapply this device to drive a LCD as a UV cell. This work demonstrates apossible scheme to function and harvest energy by an individual device.

2. Materials and experiments

2.1. Preparation of CNT-ZnO core-shell nanostructured materials

To achieve the self-powered PV device, ZnO with wide bandgap(3.3 eV), excellent thermal and chemical stability as well as opticalproperties [24], was introduced to construct the hierarchical structure

on high conductive CNT. As illustrated in Fig. 1a, ZnO as a shell waspartially deposited along the surface of the core CNT. The core-shellNWs were synthesized using an atomic layer deposition (ALD) tech-nique. First, multiwalled CNTs were grown on a silicon wafer at 1400 Kfrom a ferrocene-dissolved xylene solution (1:50, 100 c.c.). After theCNTs were synthesized, an ALD treatment was performed at 300 K todeposit ZnO onto nanotubes to achieve uniform shell layers, usingdiethylzinc and water as precursors. The as-grown CNT/ZnO core-shellstructures were then annealed under a nitrogen-rich atmosphere at1000 °C for 5min by a rapid thermal annealing system to improve thecrystalline quality. Due to the limited diffusion depth (~ 60 µm) of theALD, only the upper part of the CNT along the tube axis was coated asshown in the scanning electron microscope (SEM) image in Fig. 1b. Thediameters of the CNTs are 40–60 nm and the core-shell structures are80–100 nm. The thickness of the coated ZnO shell-layer is 30–40 nm,agreeing well with the synthesis parameters of the ALD. The mor-phology of the CNT/ZnO nanostructure is verified by transmissionelectron microscopy (TEM), as displayed in Fig. 1c. The stripe fringes(0.34 nm) within the structures represent the walls of the multi-walledCNT. The outer shell is the aggregation of ZnO nanocrystals, whosecrystallinity can be confirmed by the lattice spacing ~ 0.52 nm of ZnO(0001) [25]. The low-magnification TEM image in Fig. S1 in theSupplementary material shows that after the ALD deposition, ZnOcrystal covering on CNTs to form a rod-like morphology. By having theTEM and SEM observation, the ZnO/CNT core-shell geometry can beconfirmed.

2.2. Device fabrication

The ZnO-coated CNT PD device was fabricated in the followingprocedure. CNT/ZnO nanotubes were dispersed in isopropyl alcoholand then transferred onto a thermally oxidized Si substrate (300 nmSiO2) with prepatterned electrodes. Nanotubes were observed under aSEM and contact electrode patterns connecting nanotubes with pre-patterned electrodes were defined by electron beam lithography

Fig. 1. (a) Schematics of the core-shell nano-devices with one electrode connected to theCNT core and the other to the ZnO shell. Theinset is equivalent circuit of the core-shell na-nodevices. (b) The SEM image of the as-madeCNT/ZnO arrays after ALD deposition process.(c) High-resolution TEM image of at the CNT/ZnO junction. (d) The SEM image of the CNT/ZnO core-shell device; ZnO is highlighted withyellow false color. The inset further highlightsthe junction of the core-shell structure andbare CNT.

C.-H. Lin et al. Nano Energy 51 (2018) 294–299

295

followed by electron beam evaporation of Ti/Au (30 nm/70 nm) elec-trodes.

2.3. PD measurement

I-V properties of the CNT/ZnO core-shell nanotube PD were char-acterized using a Keithley 4200-SCS semiconductor characterizationsystem. UV light was emitted from a 325 nm UV He-Cd laser (KIMMONKoha Co., Ltd. Model: IK3552R-G). Time-resolved measurements wereachieved with the assistance of a mechanical shutter to regularly switchon/off the UV laser. Spectral response of the PD device was performedwith an EQE-R3011 spectral response system (Enli Technology). Thephotoconductive gain is calculated using the following equation:

=GΔI eP hν

//

ph

abs

where ΔIph is the difference between photo and dark current, Pabs is theabsorbed light power, e is electron charge, and hν is the photon energy,respectively.

3. Results and discussion

We first highlight the importance of the electrode contacts withrelation to the device behaviors. The Ti/Au (30 nm/70 nm) electrodeswere defined by electron beam lithography to selectively depositedonto a target region of the nanostructure. Two control experiments arepresented here (see Fig. S2): (i) both electrodes contacting to the coreCNT and (ii) both electrodes contacting to the ZnO shell. For the firstcase, since the multi-walled CNTs are typically metallic, the measuredconductance is relatively high (~ 10−6 A at 1 V) and no obvious pho-toresponse is observed, demonstrating the CNT is insensitive to light.While connecting both electrodes to the ZnO shell, the I-V curves ex-hibit symmetrical behaviors and only unsatisfactory photoresponse isobserved because the back to back Schottky contacts sufficiently pre-vent collecting photo-carriers [26]. Note that the selective illuminationto only one Schottky contact will prevent the current offset, resulting inthe boost of the photoresponse [27].

To fully activate the functionality of the core-shell nanowires, it isrequired to fabricate electrode contacts selectively to the CNT core andthe ZnO shells (Fig. 1d). According to the SEM image, 2/3 of the CNTcore coated with the ZnO shell leads to a large difference in Schottky/Ohmic contact area dominating the transport behavior of the devices.Due to the rectifying nature of the Schottky contact, the CNT/ZnO core-shell nanodevices exhibit asymmetrical I-V characteristic (Fig. 2a andan equivalent circuit is shown in the inset of Fig. 1a). The large reversedark current, as a result of the unique core-shell geometry, can be at-tributed to the large Schottky contact areas [28]. Upon illumination,the device shows evident photoresponse to the UV light (325 nm UVHe-Cd laser). Interestingly, the device performs different functionalitiesas operated at indifferent bias regions. At reverse bias region, the deviceexhibits pronounced photoresponse and works as PDs. At forward biasregion, the device possesses the PV effect that is able to harvest energyfrom the detected signals. Those features demonstrate that our CNT-ZnO core-shell nanostructures, performing both high PD and PV cap-ability, can be operated in different functionalities with or withoutexternal bias.

The photoconductive gain of a CNT-ZnO core-shell PD was in-vestigated under various-powered UV light illumination. In Fig. 2b, thephotoconductive gain at reverse bias (−2 V) is about ten times higherthan that at forward bias (2 V) regardless of the UV intensity, showingobvious bias polarity. The origin of photogain as well as its polarity isexplained as follows. Under illumination, the electron-hole pair isgenerated, and the separation of the photo-carriers is initiated by abuilt-in potential at the interface between cores and shells and the bandbending at the surface of ZnO shells. The photo-generated hole will betrapped because of the oxygen-desorption process occurring at ZnO

surface [29], leading to the circulation of electrons in circuit measuredas the photogain. When the external output is applied, the voltage isdirectly exerted upon ZnO shells in the radial direction. Specifically, forV < Voc, the built-in potential is enhanced by the external voltage andleads to an acceleration of the separated photo-carriers. This sig-nificantly reduces the carrier transition time, resulting in an amplifi-cation of photogain. For V> Voc, the built-in potential is reduced,thereby lessening the photogain of the device. To find the optimumsensitivity of the core-shell nanodevice, we have characterized the de-pendence of photogain on UV intensity, as shown in Fig. 2c. A featurecharacteristic of our device is the increase of photogain at low incidentpower, which could be an evidence of Schottky barrier forming at theinterface. Under higher light intensity, the spatial carrier separation isweakened due to significantly increased photo-carriers, lowering bar-rier height at radial Schottky junction and surface band bending; thiswould increase carrier recombination rate due to highly overlappedelectron-hole wave functions, reducing the electron lifetime andtherefore the gain [30,31]. On the other hand, barrier height at radialSchottky junction and surface band bending at low illumination in-tensities is not reduced significantly and retards the electron-hole re-combination, resulting in a high-photogain behavior, suggesting thecapability to achieve low-intensity and even a-single-photon detection.In addition, the wavelength dependence of the photogain (Fig. 2d)shows a sharp cutoff at 390 nm, confirming the superior photodetectioncapability of the device in UV wavelength region. This wavelengthdependent photogain matches the absorption spectrum of ZnO (Fig.S3), indicating that the polycrystalline ZnO dominates the light ab-sorption of the device. As a result, while CNT features polarizationdependent optical absorption [32,33], the polarization dependentphotoresponse is not obtained on the ZnO/CNT core-shell device. Fur-thermore, because both ZnO and CNT feature high thermal resistanceand long-term stability in the air [34,35], the device can work morethan three months and bear high UV light intensity up to 1000mW/cm2, which equals to 10 times of solar light intensity.

The temporal response was measured using a data acquisitionsystem (DAQ 2214, ADLINK) with a resolution of 10 μs to highlight thefast speed of ZnO/CNT core-shell UV detectors and gain insights intophoto-carrier dynamics. In Fig. 3a, the time-resolved response curvesreveal a response time of 0.48ms and a recovery time of 0.65ms at abias of − 2 V in atmosphere. The response time is defined as the timeneeded to rise to (1–1/e) of the dark current and the recovery time is(1/e) of the maximum photocurrent. The core-shell CNT/ZnO PD wasalso operated in vacuum to both demonstrate their capability for outer-space use and to further elucidate the underlying mechanisms. It is wellknown that the adsorption/desorption of the oxygen molecules is theorigin of high sensitivity of the metal-oxide sensors [36], and thus thedeficiency of the oxygen molecules will significantly degrade thefunctionality of those sensors. For example, Molina-Mendoza et al. hasreported a TiO2 nanofibres-based UV sensor featuring much betterperformance in both responsivity and response time in air than in va-cuum [37]. Generally, the response and recovery times of ZnO nano-wire (NW) PDs are several seconds to minutes, and the delayed phe-nomenon becomes even worse in vacuum due to the suppression ofoxygen readsorption [38]. Such environment-dependent factor hasbeen found in various types of ZnO-based optoelectronics, which couldbe detrimental as applying them in oxygen-deficient environments[39–41]. In Fig. 3b, it shows that our devices can still detect with fastspeed even in vacuum conditions. In contrast to the reported PDs whoseresponse/recovery times prolongs for a very long time ranging fromseveral thousand seconds to several hours [36], our device exhibits afast response time of 29ms and a recovery time of 30ms in vacuum(10−5 Torr). This improvement of photodetection speed enables manyuseful applications in outer space, for example, inter-satellite commu-nication, missile early warming and global positioning [42]. It is alsouseful components as biosensors for in-vivo applications, used to re-solute the luminescence characteristic of the biomedical agents, whose

C.-H. Lin et al. Nano Energy 51 (2018) 294–299

296

lifetime could be detectable by the improved speed. Another thingshould be noted is that both dark current and photocurrent of the de-vice in vacuum are a little higher than that in air as shown in Fig. 3.Because the oxygen molecules adsorbed on the ZnO surface can trapfree electrons and form a depletion layer near the ZnO surface, theconductivity of device slightly decreases in the oxygen-rich environ-ment.

The observed fast response/recovery speed of CNT-ZnO core-shellnanowires can be attributed to two main advantages arising from theunique geometry of the core-shell structures: (i) the thin thickness ofZnO layer and (ii) Schottky barrier built at the radial interface of thecore and shell. The shorter distance of charge transport can lead to fastcharge collection, and the strong internal field at Schottky junctioninterface help the photogenerated electron-hole pair be separatedquickly. Furthermore, CNT with extraordinary high mobility (> 105

cm2 V−1 s−1) enables the separated carrier arriving electrode at shorttime [43], which efficiently facilitates the response speed. Under illu-mination, the increase of photo-carrier density leads to a reduction ofthe barrier height, enabling the access of the photo-induced conductingpath within the depleted ZnO, and thus the conductivity is increased.Due to the thin thickness, the photo-carriers separated within the fully-depleted shell ZnO will be quickly collected by core CNT and

electrodes, giving rise to a rapid response speed. Since the process of thelight-induced Schottky barrier modulation is rather fast, the response/recovery time can be effectively reduced.

Implementation of nanodevices benefits from the employed nano-materials with improved properties as compared to their bulky coun-terparts. However, following those advantages, scaling down of thestructure also brings some detrimental effects that obstruct the nano-devices to practical uses. Here, our CNT/ZnO core-shell nanodevicebeats the challenge by fulfilling high photogain and fast detection at thesame time as well as multiple functionalities. Upon illumination, thedevice shows a pronounced PV response due to the built-in potentialfrom the Schottky junction. The short length of the nanowire device andthe large depleted region by the built-in potential enable the nanowirePD to operate at zero bias. Moreover, the I-V characteristic also exhibitsseveral notable features. First, the nanodevice shows pronouncedphotoresponse to the UV light when the external bias voltage is applied.The measured responsivity is ~ 200 A/W (at −2 V; UVintensity= 100mW/cm2), corresponding to a high gain of 800, origi-nating from the hole trapping activated by the oxygen desorption at theZnO surface and the high carrier mobility of CNT. Compared to otherrecently developed nanoscale or larger self-powered PDs (featuringresponsivities from few mA/W to 68 A/W and response times from few

Fig. 2. (a) The I-V characteristics of the CNT/ZnO core-shell PD in dark and under UV laserillumination (325 nm, 1000mW/cm2). (b) Therelationship between photoconductive gainand voltage bias with different intensities ofUV laser illumination. It is clearly shown thatphotoconductive gain at reverse bias is at leasttwo times higher than that at forward bias forlight intensities from 100mW/cm2 to1000mW/cm2. (c) Photoconductive gainmeasured as a function of excitation intensityat± 2 V bias. (d) Spectral dependence ofphotoconductive gain, demonstrating high de-tection capability of the device in UV light re-gion.

Fig. 3. (a) Time-resolved photocurrent of the CNT-ZnO core-shell device under UV laser illumination (100mW/cm2) in atmosphere. (b) Time-resolved photocurrentof the CNT-ZnO core-shell device under UV laser illumination (100mW/cm2) in vacuum.

C.-H. Lin et al. Nano Energy 51 (2018) 294–299

297

hundred µs to several s) [44–51], the device reported here shows evenbetter responsivity and comparable response time of 480 µs, demon-strating the advantages of this core-shell hierarchical structures. Sec-ondly, the device also exhibits PV behavior, where no external voltageis applied and short circuit conditions exist. Under 325 nm UV illumi-nation (~ 1000mW/cm2), the representative device yields an open-circuit voltage (Voc) of 0.85 V, a short-circuit current (Isc) around 5 nA,and a fill factor (FF) of ~25%.

Here we investigate the PV capability of the CNT/ZnO core-shelldevice as well as its applications as an UV PV cell. The representativecore-shell UV cell was ~ 3 µm long and ~ 100 nm in diameter, yieldinga Isc higher than 5 nA under 1000mW/cm2 illumination of 325 nm UVlight, as shown in the inset of Fig. 2a. Assuming that the core-shellnanostructure is a perfect cylinder, the active region, namely thejunction between core and shell can be estimated to be 9.4×10−9 cm2.Accordingly, the corresponding conversion efficiency can be calculatedto be 1.33%. This estimation does not consider the shadow effect fromnanotube. Since the shadow part of nanotube is not covered by UV il-lumination, the light active area should be smaller and the calculatedconversion efficiency is modest. Notably, the open-circuit voltage(0.85 V) is of the same level as that obtained before by Si-nanowire-based PV devices [9]. The low FF is most likely due to a consequence ofthe large series resistance in the polycrystalline ZnO shell and the re-combination lose at the interfacial defects, which is expected to beimprovable by optimizing the thickness of ZnO layer and crystal qualityof the nanostructures. This conversion efficiency and Isc, yet to be op-timized, is comparable to the reported value achieved by single nano-wire devices harvesting energy from visible solar region [52,53], in-dicating the feasibility of integrating those devices into one cell,namely, a nanoscale tandem cell.

Although the photodetection performance in the PV self-poweredmode is not as good as in external bias-powered mode (responsivity of200 A/W), the core-shell nanodevice still demonstrates superior UVdetection capability with a high responsivity of 27.5 A/W (at 0 V; UVintensity= 100mW/cm2), corresponding to a photogain of 110. Moreimportantly, in terms of power consumption, the PV mode is advanta-geous as it is self-powered. In addition, the ZnO/CNT core-shell devicecan be integrated with conventional electronics or other nanodevices toprovide energy for low power applications, in which the core-shelldevice acts as a pure PV cell. Here the PV cells and its ability to drive aLCD using UV illumination are demonstrated. One can see that the LCDcan be lighted up with an output power, where the minimum requiredpower is ~ 1 nW, corresponding to ~ 0.2 V in voltage and ~ 5 nA incurrent. A stability test of the PV cell has been performed, showing thatthe switching behavior and the fast response speed were well retainedafter each on/off cycle, as shown in Fig. 4a. In Fig. 4b, it shows a se-quence of LCD images when turn on and off the laser. As the laser wasturned on, the LCD could be immediately lit up. Since the LCD is a

capacitive device, the screen gradually faded away as the laser was off.Note that during the demonstration the device was directly connectedto the LCD screen without any output powers or measurement meters toavoid artifacts. This demonstration shows that our device can produceenough power to drive an LCD screen. Previously, the similar experi-ments have been demonstrated on triboelectric nanogenerator, whichharvest energy from mechanical motion and exhibit the same order ofoutput power [22]. The achievement of this study is to scavenge UVenergy to drive a LCD screen by an individual core-shell NW.

4. Conclusions

In summary, we have demonstrated a nanodevice with core-shellgeometry capable of UV detection and energy harvesting. A combina-tion of the advantages from the extraordinary photoresponse of ZnOand conductive CNT has greatly improved the detection speed(~ 0.5ms) without much sacrifice of the conductive gain (800). Thisdevice has proved to function with fast speed (~ 30ms) even in vacuumconditions, which is not achievable for conventional ZnO-based deviceswithout hierarchical structures. As for energy harvesting, the deviceexhibits a Voc of 0.85 V and Isc of 5 nA with a conversion efficiency of1.33%, which is capable of driving a LCD screen. We believe such core-shell schemes will provide many possibilities not only to improve theperformance of the nanodevices, but also to achieve multiple functioncapabilities.

Acknowledgements

This work was financially supported by the King AbdullahUniversity of Science and Technology (KAUST) Office of SponsoredResearch (OSR) (OSR-2016-CRG5-3005), KAUST solar center (FCC/1/3079-08-01), and KAUST baseline funding.

Competing interests

The authors declare no competing financial or non-financial inter-ests.

Appendix A. Supporting information

Supplementary data associated with this article can be found in theonline version at http://dx.doi.org/10.1016/j.nanoen.2018.06.065.

References

[1] F. Teng, K. Hu, W.X. Ouyang, X.S. Fang, Adv. Mater. 30 (2018) 1706262.[2] H.P. Wang, J.H. He, Adv. Energy Mater. 7 (2017) 1602385.[3] S. Lim, D.S. Um, M. Ha, Q.P. Zhang, Y. Lee, Y.J. Lin, Z.Y. Fan, H. Ko, Nano Res. 10

(2017) 22–36.[4] S.F. Leung, Q.P. Zhang, F. Xiu, D.L. Yu, J.C. Ho, D.D. Li, Z.Y. Fan, J. Phys. Chem.

Lett. 5 (2014) 1479–1495.[5] C.A. Lin, K.Y. Lai, W.C. Lien, J.H. He, Nanoscale 4 (2012) 6520–6526.[6] Z.Y. Fan, R. Kapadia, P.W. Leu, X.B. Zhang, Y.L. Chueh, K. Takei, K. Yu,

A. Jamshidi, A.A. Rathore, D.J. Ruebusch, M. Wu, A. Javey, Nano Lett. 10 (2010)3823–3827.

[7] Q.F. Lin, B. Hua, S.F. Leung, X.C. Duan, Z.Y. Fan, ACS Nano 7 (2013) 2725–2732.[8] C.Y. Hsu, D.H. Lien, S.Y. Lu, C.Y. Chen, C.F. Kang, Y.L. Chueh, W.K. Hsu, J.H. He,

ACS Nano 6 (2012) 6687–6692.[9] B.Z. Tian, X.L. Zheng, T.J. Kempa, Y. Fang, N.F. Yu, G.H. Yu, J.L. Huang,

C.M. Lieber, Nature 449 (2007) 885–889.[10] J.Y. Tang, Z.Y. Huo, S. Brittman, H.W. Gao, P.D. Yang, Nat. Nanotechnol. 6 (2011)

568–572.[11] C. Soci, A. Zhang, B. Xiang, S.A. Dayeh, D.P.R. Aplin, J. Park, X.Y. Bao, Y.H. Lo,

D. Wang, Nano Lett. 7 (2007) 1003–1009.[12] C.H. Lin, D.S. Tsai, T.C. Wei, D.H. Lien, J.J. Ke, C.H. Su, J.Y. Sun, Y.C. Liao, J.H. He,

ACS Nano 11 (2017) 10230–10235.[13] Z.Q. Bai, Z.W. Liu, Sci. Rep. 7 (2017) 11384.[14] C.Y. Chen, J.R.D. Retamal, I.W. Wu, D.H. Lien, M.W. Chen, Y. Ding, Y.L. Chueh,

C.I. Wu, J.H. He, ACS Nano 6 (2012) 9366–9372.[15] H.Y. Chen, K.W. Liu, L.F. Hu, A.A. Al-Ghamdi, X.S. Fang, Mater. Today 18 (2015)

493–502.[16] C.Y. Chen, M.W. Chen, C.Y. Hsu, D.H. Lien, M.J. Chen, J.H. He, IEEE J. Sel. Top.

Fig. 4. (a) A stability test of the CNT-ZnO PV cell switching on/off under UVillumination without external applied voltage bias. (b) The LCD screen was litby the CNT-ZnO core-shell device.

C.-H. Lin et al. Nano Energy 51 (2018) 294–299

298

Quantum Electron. 18 (2012) 1807–1811.[17] L.F. Hu, J. Yan, M.Y. Liao, H.J. Xiang, X.G. Gong, L.D. Zhang, X.S. Fang, Adv. Mater.

24 (2012) 2305–2309.[18] T. Rieger, D. Grutzmacher, M.I. Lepsa, Nanoscale 7 (2015) 356–364.[19] P.P. Yu, K. Hu, H.Y. Chen, L.X. Zheng, X.S. Fang, Adv. Funct. Mater. 27 (2017)

1703166.[20] X.X. Yu, H. Yin, H.X. Li, W. Zhang, H. Zhao, C. Li, M.Q. Zhu, Nano Energy 34 (2017)

155–163.[21] C.K. Jeong, K.I. Park, J.H. Son, G.T. Hwang, S.H. Lee, D.Y. Park, H.E. Lee, H.K. Lee,

M. Byun, K.J. Lee, Energy Environ. Sci. 7 (2014) 4035–4043.[22] J. Chen, G. Zhu, W.Q. Yang, Q.S. Jing, P. Bai, Y. Yang, T.C. Hou, Z.L. Wang, Adv.

Mater. 25 (2013) 6094–6099.[23] J.M. Wu, K.H. Chen, Y. Zhang, Z.L. Wang, RSC Adv. 3 (2013) 25184–25189.[24] D.S. Tsai, C.A. Lin, W.C. Lien, H.C. Chang, Y.L. Wang, J.H. He, ACS Nano 5 (2011)

7748–7753.[25] J.H. He, C.H. Ho, C.W. Wang, Y. Ding, L.J. Chen, Z.L. Wang, Cryst. Growth Des. 9

(2009) 17–19.[26] S.J. Young, L.W. Ji, S.J. Chang, Y.P. Chen, K.T. Lam, S.H. Liang, X.L. Du, Q.K. Xue,

Y.S. Sun, IET Optoelectron. 1 (2007) 135.[27] Y.F. Hu, J. Zhou, P.H. Yeh, Z. Li, T.Y. Wei, Z.L. Wang, Adv. Mater. 22 (2010)

3327–3332.[28] R.E. Burgess, H.K. Henisch, J. Electrochem. Soc. 104 (1957) (398–398).[29] A. Manor, E.A. Katz, T. Tromholt, F.C. Krebs, Adv. Energy Mater. 1 (2011) 836–843.[30] A. Zhang, S.F. You, C. Soci, Y.S. Liu, D.L. Wang, Y.H. Lo, Appl. Phys. Lett. 93 (2008)

121110.[31] A. Zhang, H. Kim, J. Cheng, Y.H. Lo, Nano Lett. 10 (2010) 2117–2120.[32] Y. Murakami, E. Einarsson, T. Edamura, S. Maruyama, Phys. Rev. Lett. 94 (2005)

087402.[33] S.C. Dhanabalan, J.S. Ponraj, H. Zhang, Q. Bao, Nanoscale 8 (2016) 6410–6434.[34] H. Beh, D. Hiller, J. Laube, S. Gutsch, M. Zacharias, J. Vac. Sci. Technol. A 35

(2017) 01B127.[35] X. Calderón-Colón, H. Geng, B. Gao, L. An, G. Cao, O. Zhou, Nanotechnology 20

(2009) 325707.[36] Q.H. Li, T. Gao, Y.G. Wang, T.H. Wang, Appl. Phys. Lett. 86 (2005) 123117.[37] A.J. Molina-Mendoza, A. Moya, R. Frisenda, S.A. Svatek, P. Gant, S. Gonzalez-Abad,

E. Antolin, N. Agrait, G. Rubio-Bollinger, D.P. De Lara, J.J. Vilatela, A. Castellanos-Gomez, J. Mater. Chem. C 4 (2016) 10707–10714.

[38] A. Bera, D. Basak, Appl. Phys. Lett. 94 (2009) 163119.[39] S.E. Ahn, H.J. Ji, K. Kim, G.T. Kim, C.H. Bae, S.M. Park, Y.K. Kim, J.S. Ha, Appl.

Phys. Lett. 90 (2007) 153106.[40] Z.Y. Fan, P.C. Chang, J.G. Lu, E.C. Walter, R.M. Penner, C.H. Lin, H.P. Lee, Appl.

Phys. Lett. 85 (2004) 6128–6130.[41] Z.Y. Fan, D.W. Wang, P.C. Chang, W.Y. Tseng, J.G. Lu, Appl. Phys. Lett. 85 (2004)

5923–5925.[42] L.X. Meng, L.Z. Zhang, Adv. Mater. Res. 299 (2011) 659–662.[43] T. Durkop, S.A. Getty, E. Cobas, M.S. Fuhrer, Nano Lett. 4 (2004) 35–39.[44] F. Bai, J. Qi, F. Li, Y. Fang, W. Han, H. Wu, Y. Zhang, Adv. Mater. Interfaces 5

(2018) 1701275.[45] H. Zhou, J. Zeng, Z. Song, C.R. Grice, C. Chen, Z. Song, D. Zhao, H. Wang, Y. Yan, J.

Phys. Chem. Lett. 9 (2018) 2043–2048.[46] B. Zhao, F. Wang, H. Chen, L. Zheng, L. Su, D. Zhao, X. Fang, Adv. Funct. Mater. 27

(2017) 1700264.[47] X. Ren, Z. Li, Z. Huang, D. Sang, H. Qiao, X. Qi, J. Li, J. Zhong, H. Zhang, Adv.

Funct. Mater. 27 (2017) 1606834.[48] H. Zhou, P. Gui, Q. Yu, J. Mei, H. Wang, G. Fang, J. Mater. Chem. C 3 (2015)

990–994.[49] J. Yu, X. Chen, Y. Wang, H. Zhou, M. Xue, Y. Xu, Z. Li, C. Ye, J. Zhang, P.A. Van

Aken, P.D. Lund, H. Wang, J. Mater. Chem. C 4 (2016) 7302–7308.[50] Z. Song, H. Zhou, P. Gui, X. Yang, R. Liu, G. Ma, H. Wang, G. Fang, J. Mater. Chem.

C 6 (2018) 5113–5121.[51] Z. Li, H. Qiao, Z. Guo, X. Ren, Z. Huang, X. Qi, S.C. Dhanabalan, J.S. Ponraj,

D. Zhang, J. Li, J. Zhao, J. Zhong, H. Zhang, Adv. Funct. Mater. 28 (2018) 1705237.[52] B. Tian, T.J. Kempa, C.M. Lieber, Chem. Soc. Rev. 38 (2009) 16–24.[53] Z.H. Yang, X.F. Li, D.Y. Lei, A.X. Shang, S.L. Wu, Nano Energy 13 (2015) 9–17.

Chun-Ho Lin is a Ph.D. student at Computer, Electrical andMathematical Sciences and Engineering (CEMSE) Division,King Abdullah University of Science & Technology(KAUST). He received the B.S. degree in physics and M.S.degree in electrical engineering from National TaiwanUniversity in 2011 and 2014, respectively. He puts his ef-forts in the development of flexible electronics using noveldevices based on nanomaterials. He is also currently in-volving in fundamental physical properties hybrid per-ovskite and their optoelectronics applications.

Hui-Chun Fu is a Ph.D. student at Computer, Electrical andMathematical Sciences and Engineering (CEMSE) Division,King Abdullah University of Science & Technology(KAUST). She received the B.S. and M.S. degree inDepartment of Physics from Tamkang University (TKU) in2006 and 2010, respectively. She puts her efforts in thesolar to hydrogen fuel and solar flow battery using Si- andGaAs-based photoelectrochemical cells with novel struc-tural design. Currently, she is involving in CO2 reduction byusing triple junction Si and triboelectric nanogenerator(TENG) devices.

Der-Hsien Lien received B.S. degree (2003) and the M.S.degree (2005) in the Department of Materials Science andEngineering from National Tsing Hua University. He re-ceived his Ph.D. degree in 2016 in the Institute ofElectronics Engineering at National Taiwan University, andcurrently works in the Electrical Engineering & ComputerSciences, UC Berkeley as a postdoctoral researcher. His re-search interests include the nanophotonics dynamics andapplications, 2D materials electronics and photonics,random access resistive memories, and flexible optoelec-tronics.

Chia-Yang Hsu received the B.S. degree in Physics fromNational Tsing Hua University, Taiwan, in 2009. He re-ceived his M.S. degree from Graduate Institute of Photonicsand Optoelectronics at the National Taiwan University,Taiwan, in 2012. His research actives include the surfaceeffects in nanostructures and electron conduction mechan-isms.

Jr-Hau He is an Associate Professor of ElectricalEngineering program at King Abdullah University ofScience & Technology (KAUST). He has been a pioneer inoptoelectronics, which reflects on his achievement ofphoton management on the light harvesting devices. He hasconducted highly interdisciplinary researches to bridgethose gaps between various research fields and betweenacademia and industry. He is a Fellow of RSC and SPIE, anda senior member of IEEE and OSA.

C.-H. Lin et al. Nano Energy 51 (2018) 294–299

299