In this study, the authors fabricated high performance core–shell nanostructured flexible photodetectors on a polyimide substrate of Kapton. For this purpose, p-type copper indium gallium selenide (CIGS) nanorod arrays (core) were coated with aluminum doped zinc oxide (AZO) films (shell) at relatively high Ar gas pressures. CIGS nanorods were prepared by glancing angle deposition (GLAD) technique using radio frequency (RF) magnetron sputtering unit at room temperature. AZO films were deposited by RF sputtering at Ar pressures of 1.0×102 mbar (high pressure sputtering) for the shell and at 3.0×103 mbar (low pressure sputtering) to create a top contact. As a comparison, the authors also fabricated conventional planar thin film devices incorporating CIGS film of similar material loading to that of CIGS nanorods. The morphological characterization was carried out by field-emission scanning electron microscope. The photocurrent measurement was conducted under 1.5 AM sun at zero electrical biasing, where CIGS devices were observed to absorb in the ultraviolet-visible-near infrared spectrum. GLAD core–shell nanorod photodetectors were shown to demonstrate enhanced photoresponse with an average photocurrent density values of 4.4, 3.2, 2.5, 3.0, and 2.5 μA/cm2 for bending angles of 0°,20°,40°,60°, and 80°, respectively. These results are significantly higher than the photocurrent of most of the flexible photodetectors reported in the literature. Moreover, our nanorod devices recovered their photoresponse after several bending experiments that indicate their enhanced mechanical durability. On the other hand, thin film devices did not show any notable photoresponse. Improved photocurrent of CIGS nanorod devices is believed to be due to their enhanced light trapping property and the reduced interelectrode distance because of the core–shell structure, which allows the efficient capture of the photo-generated carriers. In addition, enhanced mechanical durability is achieved by the GLAD nanorod microstructure on a flexible substrate. This approach can open a new strategy to boost the performance of flexible photodetectors and wearable electronics.

Flexible electronics has attracted interest from both industry and consumers due to their unique properties of being lightweight, low cost, stretchable, and wearable.1,2 A wide range of flexible devices have been known to date like photodetectors,3,4 light emitting diodes,5 pressure sensors,6 optical network communications,7 digital cameras,8,9 solar cell devices,10–12 artificial electronic skin, and biomedical sensors.13,14 However, until now, few studies were performed on the investigation of flexible inorganic semiconductor materials while most of the efforts focused on organic materials. Thus, the fabrication of flexible semiconductor materials is significantly important, which might offer us a new direction to develop a route for flexible electronics. Among these devices, photodetectors with high photoresponse, bendable and lightweight substrates, and low cost fabrication have always been an attractive research topic, due to their applications, including in military, aerospace, and medical fields, such as optical communication,15 light detections,16 digital imaging,17 energy conversion and storage,18 IR detectors,19 and wearable devices.20 Among different types of photodetectors, ultraviolet-visible-near infrared (UV-Vis-NIR) flexible detectors are playing an important role for wearable optoelectronic systems.21–23 

High-performance flexible photodetectors can be obtained by modifying the device microstructure by incorporating nanostructured semiconducting materials, which can provide mechanical flexibility while at the same time improving the optical and electrical properties. Isolated nanostructures can sustain mechanical bending or stretch more effectively due to the gaps between them. In addition, it is well known that one-dimensional (1D) nanostructures such as nanowires and nanorods exhibit superior optical absorption in comparison with thin film devices due to their enhanced light trapping originating from higher light scattering and lowered index of refraction at the nanostructured layer.24,25 In the past, several types of 1D semiconductor nanostructures have been used as an absorber material for photodetectors.26,27 Devices based on these 1D nanostructures exhibit higher performance compared to those based on traditional thin films.28 In addition to enhanced optical absorption, increased contact area and shorter distances between the electrodes within a core-shell nanostructured photodetector also allow more efficient charge carrier collection and enhanced photoresponse gain.29,30 Some researchers have used a semiconductor core with a metal shell,31 while others used semiconductors for both core and shell.32,33 With a band gap of 1.0–1.7 eV the p-type CuInxGa1–xSe2 (CIGS) is a potential candidate for high performance UV-Vis-NIR photodetectors.34,35 However, nanofabrication of compound materials such as CIGS is a challenging task. Recently, Panthani et al. successfully fabricated CIGS nanostructures using arrested precipitation technique deposition method, but the approach required a multistep processing.36 Also Singh et al. fabricated CIGS nanorods, but the fabrication approach was based on relatively complicated multistep chemical process and also required high temperatures.37 Recently, Brozak et al. successfully fabricated vertically aligned CIGS nanorods by a single-step and low-temperature method called glancing angle deposition (GLAD, also known as oblique angle deposition).34 GLAD is a simple physical vapor deposition method in which the incident flux of atoms comes to the surface of the substrate at highly oblique angles (typically more than 70°) and can be utilized in growth systems including sputtering, thermal evaporation, e-beam evaporation, and pulsed laser deposition.38–40 

In this study, we fabricated high performance flexible GLAD CIGS core-shell photodetectors followed by morphological and optical characterizations. We then performed photodetector device tests under flexing at different bending angles and compared our results to those of conventional CIGS thin film devices.

A flexible polyimide (PI) substrate with dimensions (3 × 2 cm) from DuPont was ultrasonically cleaned with methanol and deionized water then dried with nitrogen. Device fabrication was done in a multisource sputter deposition unit. All the sputter guns included NdFeB magnets. After achieving a base pressure of 3×106 mbar, a thin layer of molybdenum (Mo) film with a thickness about 50 nm was deposited by a 2 in. diameter Mo sputter target and DC sputter power of 100 W as a bottom contact on the flexible substrate in a low-pressure DC sputter reactor. A thin layer of CIGS film of about 40 nm was deposited using a quaternary 2 in. diameter CuIn0.8Ga0.2Se2 sputter target (99.99% purity from American Elements), RF sputter power of 150 W, 3×103 mbar argon (Ar) gas pressure, and a substrate rotation speed of 20 rpm. Center-to-center substrate-target distance was ∼10 cm. The substrate was azimuthally rotated around its central axis without heating. This was followed by the growth of 200 nm long vertically aligned GLAD CIGS nanorods by simply tilting the substrate to a deposition angle of 88° while keeping the other deposition conditions the same with that of the CIGS film layer. GLAD is a self-assembly simple physical growth technique. During GLAD, an incident flux of atoms comes to the substrate surface oblique angles. After a period of island formation, oblique flux causes particles to preferentially stick to the higher surface features due to the “shadowing effect” and leads to the formation of isolated and vertically aligned nanorods with the help of substrate rotation.41 To enhance the electrical contact interface of the vertically aligned, CIGS core nanorods with the charge collection transparent conduction oxide later, they were coated with a shell of aluminum-doped zinc oxide (AZO) layer by using high pressure sputtering (HIPS)42 method under 1×102 mbar Ar pressure, normal angle (deposition angle = 0°), 110 W RF power, 20 rpm substrate rotation speed, and using 2 in. diameter Al2O4(Zn) sputter target. Because of the high working gas pressure in HIPS, the mean free path of the atomic flux will decrease meanwhile their angular distribution increases, which achieve a conformal shell coating around the nanorods. HIPS AZO shell deposition was conducted for 30 min, which corresponds to an equivalent film thickness of ∼110 nm if it was deposited on a flat surface. The final stage of the capping process includes conventional low-pressure sputter (LPS) deposition of AZO at normal incidence using the same deposition parameters of the HIPS AZO layer except Ar gas pressure was reduced to 3×103 mbar and deposition time was set to 60 min. LPS AZO layer thickness is estimated to be ∼200 nm. A schematic illustration of the GLAD core–shell device is shown in Fig. 1(b). A schematic illustration of the core-shell nanorod device fabrication process is shown in Fig. 1(c). Meanwhile a CIGS thin film flexible photodetector was fabricated with same material amounts but in a classical film geometry to compare with their counterpart core–shell nanorod device as illustrated in Fig. 1(a). All the deposition parameters for thin film device were identical to those of GLAD nanorod device with the exception of a CIGS film was used instead of nanorods. However, we adjusted the volumetric amount (mass loading) of the CIGS film to be similar to the material amount of CIGS in the nanorod device. For this purpose, we measured the individual mass loadings of CIGS in nanorod and thin film device structures by conducing standalone depositions on, quartz crystal microbalance (QCM) crystals attached to the substrate holder and calculating the mass/time deposition rates by using crystal resonant frequencies before and after deposition using Sauerbrey formula.43–45 Based on our calculations, we adjusted the deposition time of the CIGS film that resulted in a film thickness of ∼150 nm. We fabricated ten devices for each nanorod and thin film geometries, where their photodetectors responses were averaged to produce the results reported in this study.

Fig. 1.

(Color online) Photodetector device schematics: (a) Thin film device construction (b) Core–shell nanorod device construction (c) Core–shell nanorod flexible photodetector fabrication process steps.

Fig. 1.

(Color online) Photodetector device schematics: (a) Thin film device construction (b) Core–shell nanorod device construction (c) Core–shell nanorod flexible photodetector fabrication process steps.

Close modal

The finalized devices were characterized by field emission scanning electron microscope (JEOL 7000F) and UV-Vis-NIR spectrometer (Shimadzu UV-3600) incorporating an integrating sphere to analyze the morphological and optical properties of the devices optical behavior. The electrical characteristics of the devices were investigated via utilizing a solar simulator under 1.5 AM using Xe light source with 150 W source power at room temperature and Keithley 2400 as a sourcemeter. The dynamic photocurrent response of the photodetector devices was investigated for both film and core–shell photodetector devices by performing time-dependent photocurrent measurements for different bending angles under zero bias (i.e., no voltage was applied) by grounding the molybdenum contact and illuminating with light from AZO contact. Finally, we investigated nanorod device's robustness under flexing by performing dynamic photocurrent measurements at different bending angles.

Figures 2(a) and 2(b) show the plane-view and cross-sectional SEM images of the thin film photodetector device on the flexible PI substrate. The average thickness of the CIGS thin film was measured as 150 nm through cross-sectional SEM analysis with 100% filling factor and zero porosity measured by the attached QCM. Figures 2(c) and 2(d) show the plane-view and cross-sectional SEM images of the CIGS (core)-AZO (shell) nanorods grown on the flexible PI substrate. Figure 2(d) shows the cross section for vertically aligned nanorod arrays of CIGS on the flexible substrate as 200 nm long grown on a thin CIGS film (40 nm) with varying diameter of nanorods (20–50 nm). From Fig. 2(d), we can see the CIGS nanorods coated with the AZO semiconductor shell by HIPS. The coated nanorods with HIPS seem denser than the straight uncoated nanorods. This is an indication that the spacing between nanorods was partially filled because the incoming AZO atomic particles have high angular distribution, which allows them to penetrate conformally in between the CIGS nanorods. In order to estimate the AZO shell thickness around the core CIGS nanorods, we compared the average diameter values obtained from cross-sectional SEM image analysis before and after the shell deposition. Based on this analysis, average shell thickness values were measured to be about 3 and 5 nm at height values 100 and 150 nm measured from the base of the nanorods. The deposition rate of the GLAD core nanorods was found 1.7 nm/min (nanorods length/deposition time) from SEM images, while for CIGS film was 8.1 nm/min. QCM was used during the deposition to measure the mass loading values for both CIGS thin film and nanorods. From the QCM data, we found that the CIGS nanorods have 85% material filing ratio per unit area, which means 15% porosity compared to the CIGS thin film.

Fig. 2.

(Color online) SEM images of flexible thin film and nanorod devices: (a) plane-view and (b) cross-sectional view of the thin film device, (c) plane-view and (d) cross-sectional view of the nanorod device on PI substrate.

Fig. 2.

(Color online) SEM images of flexible thin film and nanorod devices: (a) plane-view and (b) cross-sectional view of the thin film device, (c) plane-view and (d) cross-sectional view of the nanorod device on PI substrate.

Close modal

The optical properties of our fabricated flexible photodetector devices have been investigated by analyzing the UV-Vis-NIR absorption profiles. Figure 3 compares the optical absorption of CIGS core-shell nanorod array and thin film flexible photodetector devices. The absorption curves for both core–shell and film devices show very high absorption across the whole spectrum with greater than 95% at wavelengths in the visible range for the core-shell nanorod device. However, overall the absorption spectrum for the core–shell nanorod and thin film devices shows similar absorption values because both photodetector devices were fabricated with the same amount of each material. The core–shell photodetector exhibits complementary absorption in the UV-Vis-NIR region, making it more promising design for ultraband UV-Vis-NIR flexible photodetectors.

Fig. 3.

(Color online) UV-Vis-NIR optical absorption spectra of flexible core–shell nanorod and thin film devices.

Fig. 3.

(Color online) UV-Vis-NIR optical absorption spectra of flexible core–shell nanorod and thin film devices.

Close modal

The dynamic photoresponse of our devices was also investigated, through time-dependent photocurrent measurements using the configuration shown in Fig. 4(a) under zero biasing. We consider working with zero operating voltage because it helps better distinguishing the photocurrent, identifying the location of a junction, and it is also suitable for practical future applications. Figures 4(b) and 4(d) simply shows the time-resolved photocurrent for both devices under light illumination and zero bias.

Fig. 4.

(Color online) (a) Schematic illustration of a core–shell nanorod photodetector on PI substrate. Photocurrent density vs time for core–shell nanorod (b) and thin film (d) devices under a zero bias. (c) An enlarged portion of the nanorod device's photocurrent for a light-on-off cycle.

Fig. 4.

(Color online) (a) Schematic illustration of a core–shell nanorod photodetector on PI substrate. Photocurrent density vs time for core–shell nanorod (b) and thin film (d) devices under a zero bias. (c) An enlarged portion of the nanorod device's photocurrent for a light-on-off cycle.

Close modal

The flexible device with core–shell nanorod arrays shows a high and fast response to light compare to the flexible thin film device. We can clearly see from the curve in Fig. 4(b) that the photocurrent increases very quickly and reaches the steady state when the device is illuminated and then decreases very rapidly after the light is shuttered. On the other hand, the thin film device does not respond to light when the exposure changed. Thus, the core–shell nanorod array device has a high photocurrent and fast response to light under zero bias. The high photoresponse for the GLAD core-shell nanorod device can be explained by the following reasons. First, core–shell nanorod geometry offers enhanced light trapping.46,47 Second, the photocurrent enhancement can be attributed to charge carrier collection improvement due to the core–shell geometry which mainly shortens transition time of charge carriers.29 

Figure 4(c) shows an enlarged view of one cycle in Fig. 4(b). The rise (light turned on) and decay times (light off) of our CIGS nanorod device are about 30 and 40 ms, respectively. Such short times are much smaller than most of the other reported flexible nanorod detectors.48–50 The high surface to volume ratio of nanorods can increase defects and dangling bonds on the surface. However, an effective AZO shell coating by HIPS seems to have passivated the surface that provides the fast photodetector response observed in our nanorod devices.29 In addition, from the direction of the photocurrent, we identified that the dominant device junction is at the interface between p-CIGS nanorods and underlying Mo back contact, which indicates a Schottky type of diode behavior.

Maintaining similar photocurrent values under bending is an important parameter for flexible devices especially when it comes to their use in wearable electronic applications. The inset in Fig. 5(f) shows a picture of core–shell flexible photodetectors under bending. To investigate the photoresponse of bended devices, our flexible nanorod photodetector device were attached to a digital Vernier caliper, and dynamic photocurrent response test carried out at different bending angles θ. We measured average photocurrent densities of 4.4, 3.2, 2.5, 3.0, and 2.5 μA/cm2 for θ=0°,20°,40°,60°, and 80°, respectively, as shown in Figs. 5(a)–5(e), which indicates a small drop in photocurrent with the bending angle. The decrease in photocurrent with bending might be due to the increased gaps among the nanorods under flexing that can result in an increase in the layer resistivity and drop in the current.

Fig. 5.

(Color online) Dynamic photocurrent density profiles for nanorod arrays photodetector at different bending angles θ=0°(a),20°(b),40°(c),60°(d), and 80°(e), respectively. (f) Comparison of average photocurrent density vs bending angles for nanorod photodetector device under zero bias.

Fig. 5.

(Color online) Dynamic photocurrent density profiles for nanorod arrays photodetector at different bending angles θ=0°(a),20°(b),40°(c),60°(d), and 80°(e), respectively. (f) Comparison of average photocurrent density vs bending angles for nanorod photodetector device under zero bias.

Close modal

In addition, durability tests were carried out for the nanorod device by doing the bending test ten times for the same bending angle to check their mechanical durability. The results show stability for the photocurrent densities after each mechanical stressing of the device which indicates that nanorod device has a good mechanical durability.

In summary, we have used the GLAD and HIPS techniques to grow core (CIGS)–shell (AZO) nanorod arrays on a flexible substrate and produce a flexible photodetector. In this design, inorganic nanostructured semiconducting core offer enhanced mechanical flexibility and durability as well as superior light trapping, while the shell helps improved charge carrier collection and surface passivation. We have fabricated a core–shell CIGS nanorod photodetectors with ultrabroadband spectral detection over UV-Vis-NIR. For comparison, we also prepared conventional CIGS thin film devices with similar CIGS mass loadings to that of nanorod devices. The core–shell nanorod photodetectors were shown to demonstrate enhanced photoresponse with average photocurrent densities of 4.4, 3.2, 2.5, 3.0, and 2.5 μA/cm2 for different bending angles of 0°,20°,40°,60°, and 80°, respectively. On the other hand, thin film devices did not show any notable photoresponse. In addition, our nanorod device showed robust photoresponse after several repeated bending experiments and recovered stable photocurrent values, which indicate their enhanced mechanical durability. The presented results indicate that GLAD-HIPS method can be useful in producing high performance flexible photodetectors by using a simple and low-temperature process, which offer several opportunities, especially for wearable optoelectronic device applications.

This work was supported by NSF (Grant Nos. EPS-1003970 and 1159830). The authors would like to thank UA Little Rock Center for Integrative Nanotechnology Science for their helping with SEM and UV-Vis-NIR spectroscopy measurements.

2.
D.
Akinwande
,
N.
Petrone
, and
J.
Hone
,
Nat. Commun.
5
,
5678
(
2014
).
3.
Z.
Liu
,
G.
Chen
,
B.
Liang
,
G.
Yu
,
H.
Huang
,
D.
Chen
, and
G.
Shen
,
Opt. Express
21
,
7799
(
2013
).
4.
H.-C.
Yuan
,
J.
Shin
,
G.
Qin
,
L.
Sun
,
P.
Bhattacharya
,
M. G.
Lagally
,
G. K.
Celler
, and
Z.
Ma
,
Appl. Phys. Lett.
94
,
013102
(
2009
).
5.
6.
E.
Bosman
,
G. V.
Steenberge
,
B. V.
Hoe
,
J.
Missinne
,
J.
Vanfleteren
, and
P. V.
Daele
,
IEEE Photonics Technol. Lett.
22
,
287
(
2010
).
7.
Z.
Chen
,
W.
Ren
,
L.
Gao
,
B.
Liu
,
S.
Pei
, and
H.-M.
Cheng
,
Nat. Mater.
10
,
424
(
2011
).
8.
9.
10.
J.
Yoon
 et al,
Nat. Mater.
7
,
907
(
2008
).
11.
C.-Y.
Chang
,
Y.-J.
Cheng
,
S.-H.
Hung
,
J.-S.
Wu
,
W.-S.
Kao
,
C.-H.
Lee
, and
C.-S.
Hsu
,
Adv. Mater.
24
,
549
(
2012
).
12.
M.
Kaltenbrunner
,
M. S.
White
,
E. D.
Głowacki
,
T.
Sekitani
,
T.
Someya
,
N. S.
Sariciftci
, and
S.
Bauer
,
Nat. Commun.
3
,
770
(
2012
).
13.
W. T. S.
Huck
,
Nat. Mater.
4
,
271
(
2005
).
14.
T. Q.
Trung
and
N. E.
Lee
,
Adv. Mater.
28
,
4338
(
2016
).
15.
S.
Assefa
,
F.
Xia
, and
Y. A.
Vlasov
,
Nature
464
,
80
(
2010
).
16.
S. A.
McDonald
,
G.
Konstantatos
,
S.
Zhang
,
P. W.
Cyr
,
E. J. D.
Klem
,
L.
Levina
, and
E. H.
Sargent
,
Nat. Mater.
4
,
138
(
2005
).
17.
E.
Laux
,
C.
Genet
,
T.
Skauli
, and
T. W.
Ebbesen
,
Nat. Photonics
2
,
161
(
2008
).
18.
X.
Wang
and
G.
Shi
,
Energy Environ. Sci.
8
,
790
(
2015
).
19.
N.
Liu
,
H.
Tian
,
G.
Schwartz
,
J. B. H.
Tok
,
T.-L.
Ren
, and
Z.
Bao
,
Nano Lett.
14
,
3702
(
2014
).
20.
G.
Konstantatos
and
E. H.
Sargent
,
Nat. Nanotechnol.
5
,
391
(
2010
).
21.
S.
Chen
,
C.
Teng
,
M.
Zhang
,
Y.
Li
,
D.
Xie
, and
G.
Shi
,
Adv. Mater.
28
,
5969
(
2016
).
22.
23.
Y.
Xie
,
M.
Gong
,
T. A.
Shastry
,
J.
Lohrman
,
M. C.
Hersam
, and
S.
Ren
,
Adv. Mater.
25
,
3433
(
2013
).
24.
C.
Soci
,
A.
Zhang
,
X. Y.
Bao
,
H.
Kim
,
Y.
Lo
, and
D.
Wang
,
J. Nanosci. Nanotechnol.
10
,
1430
(
2010
).
25.
L.
Vj
 et al,
IEEE J. Sel. Top. Quantum Electron.
17
,
1002
(
2011
).
26.
T.
Zhai
,
L.
Li
,
X.
Wang
,
X.
Fang
,
Y.
Bando
, and
D.
Golberg
,
Adv. Funct. Mater.
20
,
4233
(
2010
).
27.
T.
Zhai
,
L.
Li
,
Y.
Ma
,
M.
Liao
,
X.
Wang
,
X.
Fang
,
J.
Yao
,
Y.
Bando
, and
D.
Golberg
,
Chem. Soc. Rev.
40
,
2986
(
2011
).
28.
Z.
Liu
,
J.
Xu
,
D.
Chen
, and
G.
Shen
,
Chem. Soc. Rev.
44
,
161
(
2015
).
29.
H.
Cansizoglu
,
M. F.
Cansizoglu
,
F.
Watanabe
, and
T.
Karabacak
,
ACS Appl. Mater. Interfaces
6
,
8673
(
2014
).
30.
O.
Hayden
,
R.
Agarwal
, and
C. M.
Lieber
,
Nat. Mater.
5
,
352
(
2006
).
31.
C. Y.
Hsu
,
D. H.
Lien
,
S. Y.
Lu
,
C. Y.
Chen
,
C. F.
Kang
,
Y. L.
Chueh
,
W. K.
Hsu
, and
J. H.
He
,
ACS Nano
6
,
6687
(
2012
).
32.
K.
Wang
 et al,
Appl. Phys. Lett.
96
,
123105
(
2010
).
33.
E. M.
Gallo
,
G.
Chen
,
M.
Currie
,
T.
McGuckin
,
P.
Prete
,
N.
Lovergine
,
B.
Nabet
, and
J. E.
Spanier
,
Appl. Phys. Lett.
98
,
241113
(
2011
).
34.
M.
Brozak
,
H.
Cansizoglu
, and
T.
Karabacak
,
Phys. Status Solidi RRL
11
,
1600326
(
2017
).
35.
G.
Zhang
,
S.
Finefrock
,
D.
Liang
,
G. G.
Yadav
,
H.
Yang
,
H.
Fang
, and
Y.
Wu
,
Nanoscale
3
,
2430
(
2011
).
36.
M. G.
Panthani
,
V.
Akhavan
,
B.
Goodfellow
,
J. P.
Schmidtke
,
L.
Dunn
,
A.
Dodabalapur
,
P. F.
Barbara
, and
B. A.
Korgel
,
J. Am. Chem. Soc.
130
,
16770
(
2008
).
37.
A.
Singh
,
C.
Coughlan
,
F.
Laffir
, and
K. M.
Ryan
,
ACS Nano
6
,
6977
(
2012
).
38.
K.
Robbie
and
M. J.
Brett
,
J. Vac. Sci. Technol., A
15
,
1460
(
1997
).
39.
T.
Karabacak
and
T.-M.
Lu
, “
Shadowing growth and physical self-assembly of 3D columnar structures
,” in
Handbook of Theoretical and Computational Nanotechnology
, edited by
M.
Rieth
and
W.
Schommers
(
American Scientific
,
Stevenson Ranch, CA
,
2005
), Chap. 69, p.
729
.
40.
K.
Robbie
,
M. J.
Brett
, and
A.
Lakhtakia
,
Nature
384
,
616
(
1996
).
41.
A.
Barranco
,
A.
Borras
,
A. R.
Gonzalez-Elipe
, and
A.
Palmero
,
Prog. Mater. Sci.
76
,
59
(
2016
).
42.
F.
Keles
,
H.
Cansizoglu
,
E. O.
Badraddin
,
M. P.
Brozak
,
F.
Watanabe
, and
T.
Karabacak
,
Mater. Res. Express
3
,
105028
(
2016
).
43.
M. T.
Demirkan
,
L.
Trahey
, and
T.
Karabacak
,
Thin Solid Films
600
,
126
(
2016
).
44.
T.
Ozkan
,
M. T.
Demirkan
,
K. A.
Walsh
,
T.
Karabacak
, and
A. A.
Polycarpou
,
J. Mater. Res.
31
,
2011
(
2016
).
45.
M. T.
Demirkan
,
L.
Trahey
, and
T.
Karabacak
,
J. Power Sources
273
,
52
(
2015
).
46.
M. F.
Cansizoglu
,
R.
Engelken
,
H.-W.
Seo
, and
T.
Karabacak
,
ACS Nano
4
,
733
(
2010
).
47.
H.
Cansizoglu
,
M. F.
Cansizoglu
,
M.
Finckenor
, and
T.
Karabacak
,
Adv. Opt. Mater.
1
,
158
(
2013
).
48.
S. R.
Tamalampudi
,
Y.-Y.
Lu
,
R.
Kumar
,
U. R.
Sankar
,
C.-D.
Liao
,
B. K.
Moorthy
,
C.-H.
Cheng
,
F. C.
Chou
, and
Y.-T.
Chen
,
Nano Lett.
14
,
2800
(
2014
).
49.
L.
Li
,
Z.
Lou
, and
G.
Shen
,
ACS Appl. Mater. Interfaces
7
,
23507
(
2015
).
50.
T.
Zhai
 et al,
ACS Nano
4
,
1596
(
2010
).