- Email: [email protected]
Wafer-scale high-quality Ag thin film using a ZnO buffer layer for plasmonic applications
Applied Surface Science 512 (2020) 145705
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Full Length Article
Wafer-scale high-quality Ag thin film using a ZnO buffer layer for plasmonic applications
T
Bo-Gwang Junga, Miyeon Cheonb, Su Jae Kimb, Alexander Gliserinc, Soo Hoon Chewc, ⁎ ⁎ ⁎ Chae Ryong Chod, Seong-Gon Kime, Young Hee Leef,g, , Seungchul Kima,c, , Se-Young Jeonga,c, a
Department of Cogno-mechatronics Engineering, Pusan National University, Busan 46241, Republic of Korea Crystal Bank Research Institute, Pusan National University, Busan 46241, Republic of Korea c Department of Optics and Mechatronics Engineering, Pusan National University, Busan 46241, Republic of Korea d Department of Nanoenergy Engineering, Pusan National University, Busan 46241, Republic of Korea e Department of Physics and Astronomy, Mississippi State University, Mississippi State, MS 39762, USA f Center for Integrated Nanostructure Physics, Institute for Basic Science, Sungkyunkwan University Suwon 16419, Republic of Korea g Department of Energy Science, Sungkyunkwan University, Suwon 16419, Republic of Korea b
A R T I C LE I N FO
A B S T R A C T
Keywords: Ag thin film EADM Reflectivity Surface plasmon ZnO buffer layer
Realizing laterally continuous, ultraflat silver (Ag) single-crystal films is a significant technological challenge. Ag thin film grown on various hetero-substrates has been used in numerous applications, due to its superior electrical and optical properties. To exploit these properties without degradation and apply these films to highprecision patterning, surface plasmonics, and so on, a high-quality thin film having an ultraflat surface and few grain boundaries is needed. A zinc oxide (ZnO) buffer layer can be used to facilitate the growth of a singlecrystalline Ag thin film on a sapphire (Al2O3) substrate. ZnO films deposited on Al2O3 substrates have grain boundaries; however, Ag films grown on ZnO are nearly grain-free and close to single-crystalline quality. This can be explained by the exceptionally small extended atomic distance mismatch, of ~0.08%, between Ag and Al2O3, in which the ninth Ag atom and the eighth Al atom are matched in terms of coherence and periodicity. A modified radio frequency sputtering system with a single-crystal Ag target enabled wafer-scale growth of ultraflat, grain-free Ag films. The proposed approach using a ZnO buffer provides a new method for fabricating Ag films with high adhesion, anti-oxidative stability, and superior optical properties, and allows for easy nanopatterning.
1. Introduction Noble metal films, such as gold (Au) and silver (Ag) films, have attracted considerable interest in optical nano-circuitry, next-generation photovoltaic devices, and surface plasmonic resonance [1–4]. Agcoated mirrors are ideal for observing a wide range of light wavelengths, having high reflectivity and low emissivity that is especially useful in infrared applications for astronomy [5]. Because electromagnetic coupling between light and free electrons at metal surfaces is important in surface plasmons, surface roughness and other inhomogeneities of the metal that limit surface-plasmon propagation are a major barrier to efficient plasmonic devices [6–9]. In a study comparing the optical response of two plasmonic nanoantennae, distinct differences between two structures, where one had an idealized geometry and the other a realistic geometry, resulted in
approximately 1.4-fold differences in the near-field properties, polarization, and spectral behavior [10]. Deep subwavelength integration of high-definition plasmonic nanostructures, which is of key importance in the development of future optical nanocircuitry, depends on commonly used thermally evaporated Au layers. Evaporated Au layers tend to have structural imperfections that appear in individual circuit elements, resulting in a drastic reduction in the yield of integrated nanocircuits. Chemically grown single-crystalline Au flakes have been shown to provide an ideal basis for focused ion beam (FIB) milling and other topdown nanofabrication techniques, to produce high-definition, ultrasmooth Au nanostructures with superior optical properties [11]. Although the fabrication of single-crystalline metallic nanostructures using chemical processing methods has been reported [12,13], a physical technique, such as evaporation or sputtering, is preferred, because chemical processing methods make it difficult to control uniformity due
⁎ Corresponding authors at: Center for Integrated Nanostructure Physics, Institute for Basic Science, Sungkyunkwan University Suwon 16419, Republic of Korea (Y.H. Lee). Department of Cogno-mechatronics Engineering, Pusan National University, Busan 46241, Republic of Korea (S. Kim and S.-Y. Jeong). E-mail addresses: [email protected] (Y.H. Lee), [email protected] (S. Kim), [email protected] (S.-Y. Jeong).
https://doi.org/10.1016/j.apsusc.2020.145705 Received 20 December 2019; Received in revised form 30 January 2020; Accepted 8 February 2020 Available online 08 February 2020 0169-4332/ © 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Applied Surface Science 512 (2020) 145705
B.-G. Jung, et al.
white plate (barium sulfate). The incident angle of the beam was set at 5°.
to the metallic nanoclusters that form and subsequently complicate the synthesis process [14,15]. Extremely flat, single-crystalline Ag films with a root mean square (RMS) roughness of 0.7 nm and grain size of several hundred nanometers have been obtained using a standard direct current magnetron sputtering system; the films possessed several key properties necessary for plasmonics applications [7]. A single-crystalline Ag film grown on a SiO2 substrate was used to obtain a Ag nanopillar structure by FIB milling; [16,17] observation of the crystal orientation of the Ag films using an electron backscatter diffraction (EBSD) method confirmed grain growth, achieved by adjusting the film deposition conditions. However, the grain size under the optimum conditions was no more than ~200 nm. Thus, growth of grain-boundary-free single-crystalline metal films and replacement of Au film with copper or Ag single-crystalline film remains challenging. Sputtering is expected to provide an innovative solution to the aforementioned problems and enable waferscale production. In our previous studies, we introduced a method to obtain highquality wafer-scale grain-boundary-free single-crystalline copper thin films on a sapphire (Al2O3) substrate using a sputtering method [18,19]. However, Ag films grown by the same method showed unexpectedly poor adhesion to the Al2O3 substrate due to the large atomic distance mismatch (ADM), which led to film delamination from the substrate. In this letter, we introduce the growth of ultraflat, wafer-scale single-crystalline Ag thin films by incorporating a zinc oxide (ZnO) buffer layer on the Al2O3 substrate. We describe how the Ag atoms overcome the large ADM, with the substrate using the ZnO buffer layer to relieve the large strain induced by heteroepitaxy, thus resulting in single-crystalline film with an ultraflat surface and high crystallinity.
3. Results and discussion Ag has a large ADM of 4.9% or 5.1% with respect to the Al2O3 substrate, varying depending on the growth direction. Despite efforts to optimize film growth conditions, Ag film on the Al2O3 (0 0 6) substrate is still of poor quality, and separates easily from the substrate. The quality of Ag films grown on a Al2O3 (0 0 6) substrate with previously reported conventional methods was not sufficient for plasmonic applications. The typical crystalline properties of our Ag film grown on Al2O3 are presented in Fig. S1 (Supporting Information). The EBSD orientation map (Fig. S1a, Supporting Information), its inverse pole figure (IPF) (Fig. S1b, Supporting Information) [3], pole figure (PF) with respect to the normal direction, i.e., ∥ [1 1 1] of Ag (Fig. S1c, Supporting Information), and X-ray diffraction (XRD) pattern (Fig. S1f, Supporting Information) show the polycrystalline nature of the Ag thin film on Al2O3, despite the use of a modified sputtering method. Every grain in the polycrystalline Ag was aligned in an arbitrary direction, and the grain size was, on average, several tens of nanometers. Scanning electron microscopy (SEM) images obtained under low (Fig. S1d, Supporting Information) and high magnification (Fig. S1e, Supporting Information) confirmed the EBSD results, i.e., that the grains of the Ag film are deposited in arbitrary directions with a grain size of several tens of nanometers. The surface roughness obtained from atomic force microscopy (AFM) measurements (Fig. S1g, Supporting Information) indicated an RMS value of 5.9 nm, which increased over time due to further oxidation in a normal laboratory environment. Based on these results, we concluded that the Al2O3 (0 0 6) substrate was not satisfactory for Ag film growth, and we proceeded to identify a reasonable buffer layer to reduce lattice mismatch and release the strain induced by the large ADM. Given that we could not improve the crystallinity of the Ag film using a Si substrate, we chose Al2O3 as the substrate material. ZnO was determined to be the best candidate as a buffer layer for the Al2O3 substrate, as it shows a good structural match with Al2O3 and is chemically stable. ZnO film has been grown on Al2O3 (0 0 6) substrates for lightemitting diode (LED) and dilute magnetic semiconductor applications, and is well known to be compatible with Al2O3 (0 0 6) substrates [20,21]. The schematic structure of an Ag thin film, on a ZnO buffer layer on a Al2O3 (0 0 6) substrate, and an actual photograph of a 50-nmthick Ag thin film grown on a ZnO buffer layer are shown in Fig. 1a. XRD images were taken of the areas marked by the colored squares in Fig. 1a; the color of the curves in Fig. 1b corresponds to the area marked by the same color in Fig. 1a. The similar XRD results among the different areas prove the wafer-scale homogeneity of the Ag thin film on the ZnO buffer layer. The peaks at 34.4°, 38.2°, and 41.7° correspond to those of ZnO (0 0 2), Ag(1 1 1), and Al2O3 (0 0 6), respectively. An AFM image of the Ag film (Fig. 1c) shows a very flat surface in the vertical scale of a full height of 20 nm; for the 10 µm × 10 µm area; the RMS surface roughness value was 0.73 nm, slightly larger than the RMS roughness (~0.26 nm) of the Al2O3 substrate. The right upper inset shows an AFM image of a 2 µm × 2 µm area with a higher contrast in the vertical scale of 5 nm, which was acquired to better resolve the surface roughness. The Al2O3 substrate-ZnO buffer layer-Ag film interfaces were characterized using cross-sectional high-resolution transmission electron microscopy (HR-TEM). The cross-sectional HRTEM images with different resolutions reveal epitaxial growth of the Ag film (Fig. 1d–f). The fast Fourier transform (FFT) patterns of the areas in Fig. 1f marked with rectangular boxes are shown in Fig. 1g–i. The orientation of the Al2O3 substrate parallel to the electron beam direction is 〈110〉, whereas the orientations of the ZnO buffer layer and the Ag film are 〈21¯0〉 and 〈110〉, respectively. The (0 0 6) planes of Al2O3, ZnO
2. Materials and methods Ag thin films were fabricated using a modified radio frequency sputtering method. The single-crystal Ag target used in the sputtering process was prepared as 2-inch-diameter disks derived from singlecrystal Ag ingots grown by the Czochralski method, supported by the Crystal Bank of Pusan National University (Busan, Korea). Special care was taken to reduce the instability of the sputtering system caused by the electrical power supply and mechanical vibrations. Ag thin films were grown on 5 nm thick ZnO thin film sputtered on a 2-inch Al2O3 (0001) substrate under various conditions. The base pressure was ~10−6 Torr, and the working pressure, of 10−3, was adjusted by highpurity Ar gas (99.999%). The growth rate was 1 Å/s. The θ-2θ XRD measurements were performed using an Empyrean Series 2 instrument (PANalytical B.V., Almelo, The Netherlands) equipped with a Cu-Kα source (40 kV, 30 mA). Data were collected within the range of 20° < 2θ < 80°, with a step size of 0.0167°. AFM measurements were carried out using an XE-100 instrument (Park Systems, Inc., Suwon, Korea). AFM images of the Ag thin films (10 × 10 µm2) were obtained in non-contact mode (cantilever: PPPNCHR). The SEM and EBSD images were obtained using a SUPRA40 VP system (Carl Zeiss AG, Oberkochen, Germany). The PF represents the spatial orientation of grains and the orientation distribution of crystallographic lattice planes in stereographic projection. The IPF shows the pole that is parallel to a given thin film direction. The IPF represents only a unit triangle, as opposed to a full circle, due to crystal symmetry. The IPF map combines the locally detected orientation with the crystallographic data. The color coding is in accordance with the reduced size of the IPF. The electrical resistivity of the Ag thin film was measured using the van der Pauw method (HMS-3000; ECOPIA, Anyang, Korea) and the diffuse reflectance was measured using an ultraviolet-visible spectrometer (V-770; Jasco Inc., Easton, MD, USA) equipped with a 60-mm Φ integrating sphere. The reflectivity was measured over the range 180–1000 nm. The baseline correction procedure used a reference 2
Applied Surface Science 512 (2020) 145705
B.-G. Jung, et al.
Fig. 1. Structural details of a single-crystalline silver (Ag) thin film (thickness: 50 nm) grown on a 5 nm zinc oxide (ZnO) buffer film. (a) Schematic diagram and photographic images of an Ag film on ZnO/sapphire (Al2O3). (b) X-ray diffraction (XRD) patterns corresponding to the areas marked in (a). (c) Atomic force microscopy (AFM) images of a single-crystalline Ag thin film on the ZnO buffer layer with the vertical scale of the full height of 20 nm (inset with 5 nm). (d–f) Crosssectional high-resolution transmission electron microscopy (HR-TEM) images, under different magnifications, of a single-crystalline Ag thin film on a ZnO buffer layer. (g–i) Fast Fourier transform (FFT) patterns of Al2O3, ZnO, and Ag, corresponding to the areas in the HR-TEM image (f) marked by rectangular boxes.
exhibited a dark gray color. The XRD and SEM results of the Ag thin films grown on the ZnO buffer layer deposited at 500 °C are shown in Fig. 2c and d, respectively. The thickness of Ag films was adjusted to ~100 nm during the optimization process. In the XRD data presented at the linear scale (Fig. 2c), only three peaks appeared, corresponding to the ZnO (0 0 2) buffer layer, Ag (1 1 1) film, and Al2O3 (0 0 6) substrate at 34.4°, 38.2°, and 41.7°, respectively. The intensity of XRD for each sample was normalized with respect to the Al2O3 (0 0 6) intensity, to estimate the crystallinity of the Ag (1 1 1) peak. The intensity of the Ag (1 1 1) peak for all samples was at least three-fold higher than the peak intensity of the Al2O3(0 0 6) substrate itself, whereas the Ag (1 1 1) peak of the polycrystalline Ag film grown directly on the Al2O3(0 0 6) substrate (Fig. S1f, Supporting Information) was almost hidden by the Al2O3(0 0 6) substrate peak, and was not observed at all at the linear scale. Considering the highest peak intensity and smallest full width half maximum of the Ag (1 1 1) (0.18°) peak, the optimum condition for Ag film growth on ZnO was taken to be 400 °C in this study. Fig. 2d shows SEM images of the samples grown at 200 °C, 300 °C, 400 °C, and 500 °C. Considering that the ZnO buffer layer was aligned with the [1 0 0] direction, but was still polycrystalline, the smooth surfaces of all four samples observed in low-magnification SEM images (1-μm scale; insets in Fig. 2d) were surprising. A clear grain boundary could not be identified for the sample grown at 400 °C, even under high magnification. This suggests that the Ag film was nearly single-crystalline. We examined the RMS values of Ag films as a function of growth temperature (Fig. S2b, Supporting Information), where the RMS value was 1.50 nm for the sample grown at 200 °C and 0.73 nm for that grown at 400 °C. The right panel of the AFM images indicates the line profiles across the AFM surface images, and the RMS surface roughness values calculated for a 10 µm × 10 µm area of the thin film. These images further confirmed the single-crystal-like growth structure on the polycrystalline substrate, for which heteroepitaxy is very unusual. To understand the growth mechanism, it is necessary to investigate the crystallographic structure of the grown Ag film in detail. To determine the crystal orientation in the plane, and to understand the grain growth process, we examined coincident site lattice (CSL) boundary
(0 0 1), and Ag (1 1 0) are normal with respect to the interfaces; the respective normal values for these directions, 〈21¯0〉 of Al2O3, 〈010〉 of ZnO and 〈11¯0〉 of Ag, are also parallel to each other. Thus, HR-TEM analysis supports the observed growth in the 〈111〉 direction of the Ag thin film on a Al2O3 substrate. We attempted to grow the ZnO buffer film on Al2O3 substrates under various growth temperatures, ranging from 200 °C to 500 °C, to determine the optimum growth conditions based on XRD and SEM measurements. The films were adjusted to have the same thickness, of ~100 nm. The XRD results for the ZnO buffer films are presented on a linear scale; the intensities of ZnO peaks were normalized with respect to the Al2O3 (0 0 6) intensity in Fig. 2a. All samples showed clear ZnO (0 0 2) and Al2O3 (0 0 6) peaks at 34.4° and 41.7°, respectively, which indicates that the ZnO films were well aligned in the [0 0 1] direction along the Al2O3 (0 0 6) plane. The two samples grown at 400 °C and 500 °C showed a single phase and high crystallinity. However, SEM revealed larger grains in the sample grown at 500 °C (Fig. 2b). Lowmagnification SEM images of the ZnO (insets in Fig. 2b) confirmed that the grown ZnO films were of good quality and uniform, regardless of the growth temperature over the range 200–500 °C. High-magnification SEM images clearly revealed grain growth (20–50 nm in size), where grain size increased with growth temperature up to 500 °C. At temperatures higher than 500 °C, the film quality of ZnO deteriorated (not shown here). Based on these results, we limited the growth temperature of the ZnO film to below 500 °C. Notably, the ZnO film grown at 500 °C showed the best crystallinity and largest grain size among the film samples; additionally, the ZnO films grown at different temperatures showed no clear difference in RMS roughness (Fig. S2a, Supporting Information). Thus, 500 °C was chosen as the optimal growth temperature of the ZnO buffer layer with a thickness of 5 nm for Ag film growth; all other growth conditions, such as the initial pressure, working pressure, Ar flow, and film thickness were held constant during Ag film growth. Once again, we varied the growth temperature of the Ag film over the range of 200–500 °C to identify the optimal Ag film growth conditions. Notably, Ag film grown above 600 °C showed oxidization and 3
Applied Surface Science 512 (2020) 145705
B.-G. Jung, et al.
Fig. 2. Optimization of the growth temperature of a ZnO buffer film and Ag film. (a, b) XRD (a) and scanning electron microscopy (SEM) images (b) of ZnO films. ZnO buffer films deposited on Al2O3 (0 0 6) substrates; the optimum condition was 500 °C. (c, d) XRD (c) and SEM images (d) of Ag thin films deposited on a 5 nm ZnO buffer layer. The single-crystalline Ag thin films were grown on a Al2O3 substrate using a ZnO buffer layer under the optimum condition of 400 °C.
line maps of Ag thin films using EBSD. Crystallographic characterization of Ag thin films was performed. The CSL boundary lines in the EBSD maps of Ag thin films, grown on a Al2O3 substrate using a ZnO buffer layer, showed the changes in film crystallinity with growth temperature (Fig. 3a); the blue color corresponds to the crystal orientation along the 〈111〉 direction, and the CSL boundary lines denote the boundaries between areas with different orientations. The grain size was estimated based on the CSL boundary lines; the entire measured area of all samples was aligned, or nearly aligned, with the 〈111〉 direction, in contrast to the polycrystalline sample (Fig. S1a, Supporting Information). The IPFs corresponding to the EBSD orientation maps shown in Fig. 3a show the degree to which each grain is aligned with the (1 1 1) plane (Fig. 3b); the dispersion of the spots in the area near (1 1 1) was caused by the tilting of two twins with different stacking orders.[19] The CSL boundary lines in the EBSD maps showed that the sample grown at 200 °C had relatively smaller grains. The grains of samples grown at 300 °C and 500 °C were larger, but showed a slight difference in degree of tilt from the (1 1 1) plane; thus, the spots in IPF show a wide dispersion. The six-fold symmetry in the EBSD {1 0 0} PF with respect to the normal direction, i.e., ∥ [1 1 1] (Fig. 3c), arises from the two grains rotated by 60° with an ABCABC⋯ and ACBACB⋯ stacking order [19]. Although the spots in PF (Fig. 3c, insets) were located within a small area, the spots in the IPF were widely dispersed or separated into two clusters, as each of the three originates from two grain types of different stacking orders. The IPF under the 400 °C condition was better aligned in the (1 1 1) direction; two grain types (same color on the EBSD map) were well aligned in the (1 1 1) plane. The CSL boundary lines in the EBSD maps are represented in red, and the rotation angle between the two grains ranged between 59°and 60°. Parts of the red CSL boundary lines were larger than 99.5% for the sample grown at 400 °C (see the rotation angles below the corresponding EBSD
map in Fig. 3), which implies that two grain types were typically rotated by nearly 60° with respect to the other. Ideally, the {1 0 0} PF of the cubic structure with respect to the normal direction, i.e., ∥ [1 1 1], should have three-fold symmetry; however, the six-fold symmetry of the φ-scan XRD pattern of the Ag (1 1 1) plane (Fig. 3d) implies 60° rotation of two types of grains in the thin film. The peaks marked by ▲ and ▽ on the XRD φ-scan represent the two grain types; the equivalent markers in the EBSD map obtained under the 400 °C condition indicate the cause of the six-fold symmetry in the {1 0 0} PF in Fig. 3c, and the XRD peaks in Fig. 3d, respectively. The formation of the two kinds of grains rotated by 60° was attributed to the two stacking orders, ABC… and ACB…, in the face-centered cubic (1 1 1) plane (Fig. 3e). These two grains are connected by a hidden incoherent twin boundary (HITB) that does not permit vacant atoms; thus, the boundary assists overlayer growth without defects. The mechanism of HITB formation, and the growth of the grains to form a single grain, were first discussed in a study of copper single-crystal thin films; [19] interestingly, the Ag thin film is grown on the ZnO buffer layer according to the same mechanism by which copper thin film is grown on Al2O3. Ag film growth on the ZnO buffer layer of an Al2O3 substrate involved two types of heteroepitaxy. The ZnO film worked well as a buffer; however, it retained a polycrystalline form and, thus was not single-crystal in nature to the same extent as the Al2O3 substrate. Furthermore, the large ADM between Ag and ZnO makes it difficult to explain how Ag thin film of single-crystalline quality can grow on the polycrystalline ZnO buffer layer. Thus, we used an EADM model to determine whether there was extended periodicity of the lattice match between the two materials. The structural relations among the substrate, buffer, and final growth material are illustrated by the superposition of atomic layers in Fig. 4a. According to the Moire pattern observed with lattice superposition, overlayer growth can be predicted. 4
Applied Surface Science 512 (2020) 145705
B.-G. Jung, et al.
Fig. 3. Crystallographic orientations of Ag films grown on a ZnO buffer film under various temperatures: (a) coincident site lattice (CSL) boundary line maps of electron backscatter diffraction (EBSD) measurements of Ag thin films. All scale bars are 1 μm; (b, c) inverse pole figure (IPF) corresponding to the EBSD orientation maps of (a) (b) and EBSD [4] pole figure (PF) with respect to the normal direction, i.e., ∥ Ag [1 1 1] (c). (d, e) φ-scan XRD pattern of the Ag (1 1 1) plane (d) and an atomic model showing two possible stacking configurations on the hexagonal basal plane, in which a grain is rotated by 60˚ (e).
the epilayer [23]. The lattice mismatch calculated between ZnO and Al2O3 was ~18%, as shown in Fig. 4b. The overlapping crystal structures of ZnO and Al2O3 indicate that these two materials have extended periodicity; however, the respective EADMs were 0.51% and 1.5%, which are still relatively large. Based on this result, it is inevitable that the ZnO buffer layer grown on Al2O3 has grain boundaries, even under optimized conditions (Fig. 3b). On the other hand, Ag shows remarkable similarity to ZnO; while the ADM was ~11% for both materials, the EADM was ~0.08%. Because nine times the Ag-Ag distance of Ag is approximately equal to eight times the ZnZn distance of ZnO, I and I′ are 9 and 8, respectively, and the larger of the period lengths in two directions is about 5.0 nm, which is much smaller than that between Ag and Al2O3. The small EADM, I and I′ values promote ultraflat growth of Ag film on ZnO with single-crystalline quality. All results of ADM and EADM are summarized in Table 1. The electrical conductivity and optical reflectivity are representative physical quantities sensitive to grain boundary formation. The scattering of electrons at grain boundaries and rough surfaces is the main cause of resistivity and deterioration of electric properties. The reduction in resistivity achieved by eliminating grain boundaries amounts to about 13% in bulk copper and Ag substrates [26,27]. The
Generally, oxides such as Al2O3 are ionic crystals, in which small-size cations, such as A13+, with larger valences easily attract and retain oxygen ions. The surface of Al2O3 consists only of oxygen ions [22]. Thus, the lattice mismatch can be calculated with respect to the change in Ag-O atomic distances between Ag and Al2O3. The O-O distance of ¯ 10] Al2O3 in the [1 0 0] direction was 2.753 Å, and that of Ag in the [1 direction was 2.889 Å. The degree of ADM was 4.9%, with 20:21 being the smallest integral ratio (2.889:2.753). Twenty times the Ag-Ag distance of Ag is nearly equal to twenty-one times that of the O-O distance of Al2O3, as shown in Fig. 4a. The EADM for large lattice mismatch epitaxial growth can be calculated as follows: EADM = Id − I 'd' , where I and I′ are integers and d and d′ are atomic I 'd' distances of the epilayer and substrate, respectively. I and I′ are determined as follows: d:d′ ~ I:I′, where I:I′ is the smallest integral ratio for d:d′. The difference between I and I′ introduces a periodic edge-type dislocation [23–25]. Although the EADM for Ag and Al2O3 is small (0.08%), I and I′ are 20 and 21, respectively. This period length in Ag in the [1¯10 ] vertical direction is about 5.78 nm, and that in the [112¯ ] horizontal direction is about 10.01 nm, which is sufficiently long for other defects to potentially appear. The lowest mismatch limit that can be applied to EADM is not fixed, and depends on the deformation potential of the epilayer, as well as on the nature and number of defects in
Fig. 4. Extended atomic distance mismatch between the film and substrate lattice: superposition of (a) Ag in Ag (1 1 1) on O in Al2O3 (0 0 1), (b) Zn in ZnO (0 0 2) on O in Al2O3 (0 0 1), and (c) Ag in Ag (1 1 1) on O in ZnO (0 0 2). 5
Applied Surface Science 512 (2020) 145705
B.-G. Jung, et al.
by the absolute decrease in electron scattering seen at the grain boundaries and on rough surfaces. The reflectivity of the single-crystalline 50-nm-thick Ag thin film was compared to that of a commercial polycrystalline 150-nm-thick Ag thin film (G-mek, Inc., Gyeonggi, Korea). The single-crystalline Ag thin film showed clearly superior reflectivity over the measured wavelength range. To determine the chemical stability of the single-crystalline Ag thin film, its reflectivity was measured again 6 months later and similar results were obtained [Fig. 5(b) and Supplementary Fig. S3]. The fact that the reflectivity of our Ag film was maintained even after several months indicates the chemical stability of the film; notably, a protective layer for commercial Ag mirrors is mandatory. Surface plasmon resonance (SPR) curves were obtained for singleand polycrystalline Ag films using a coherent 632.8-nm light source (Fig. 5c). In the single-crystalline Ag film, a deeper and sharper resonance curve was obtained compared to that for the polycrystalline Ag film, with the same film thickness. The position and depth of the dip in SPR well-matched that obtained with numerical simulations of a singlecrystalline Ag film (Supplementary Fig. S4). This implies that the singlecrystalline film should have a higher real permittivity and lower imaginary permittivity. The width of the plasmonic dip depends on the imaginary part of the dielectric constant of the metal. Such an SPR angle mismatch between single- and polycrystalline films may originate from the ZnO interlayer film below the single-crystalline Ag film. However, numerical simulations showed that the ZnO buffer layer did not affect the SPR curve (Supplementary Fig. S4). The chemical stability of the single- and polycrystalline Ag films was evaluated based on the SPR curve, after exposing the samples to ambient conditions for 2 months (Fig. 5d and e). The shift in SPR curve of the single-crystalline
Table 1 Atomic distance mismatch (ADM) and extended atomic distance mismatch (EADM). Material
Atomic distance (Å)
ADM
I
I′
EADM
Ag(1 1 1)/Al2O3(0 0 6)
D1 = 2.889 D2*=2.753 D′1 = 5.004 D′2 = 4.761 D3 = 3.248 D2 = 2.753 D′3 = 5.625 D′2 = 4.761 D1 = 2.889 D3 = 3.248 D′1 = 5.004 D′3 = 5.625
4.9% 5.1% 18.0% 18.1% 11.1% 11.0%
20 20 6 5 9 9
21 21 7 6 8 8
0.08% 0.1% 0.51% 1.5% 0.07% 0.08%
ZnO(0 0 2)/Al2O3(0 0 6) Ag(1 1 1)/ZnO(0 0 2)
D1, D2, and D3 are the vertical atomic distances of Ag-Ag, O-O in Al2O3, Zn(O)Zn(O) respectively. D′1, D′2, and D′3 are the horizontal atomic distances of AgAg, O-O in Al2O3, Zn(O)-Zn(O), respectively. * The mean distance of vertical direction oxygen atoms at Al2O3 was calculated as (2.868 × 2 + 2.524)/ 3 = 2.753 (Å).
minimum resistivity that can be achieved in a copper single-crystal film was reported as 1.61 μΩ·cm, which is 7% lower than the resistivity of Cu bulk according to the International Annealed Copper Standard (IACS: 1.7241 μΩ·cm), and 6% higher than that of single-crystal Cu bulk (1.52 μΩ·cm at 293 K) [18]. For our Ag thin films deposited on ZnO, no clear grain boundaries were observed. The film surface was ultraflat, with an RMS roughness nearly a quarter of that of polycrystalline Ag films (Fig. S2a, Supporting Information). In this study, the measured resistivity of the single-crystalline Ag thin films (thickness: 50 nm) was < 3 µΩ·cm; the resistivity of the samples grown under 400 °C and 500 °C was ~1.87 µΩ·cm, which is close to that of Ag bulk (1.6 µΩ·cm); the typical resistivity of polycrystalline Ag is 2.61 µΩ·cm [28] (Fig. 5a). These results suggest that the mean free path of electrons was increased
Fig. 5. Physical properties of Ag single-crystalline film grown using a ZnO buffer layer. (a) Resistivity of the Ag films as a function of growth temperature. (b) Reflectivity of the Ag thin film as a function of wavelength compared to a commercial Ag thin film. (c) Surface plasmon resonance (SPR) reflectivity of the singlecrystalline Ag film with a ZnO buffer layer of 5 nm thickness as a function of incident angle just after deposition compared to a polycrystalline Ag thin film without a ZnO buffer layer. (d, e) SPR reflectivity of a polycrystalline Ag thin film without a ZnO buffer layer (d) and a single-crystalline Ag film after 2 months (e). 6
Applied Surface Science 512 (2020) 145705
B.-G. Jung, et al.
Ag film was negligible; however, the SPR curve of the polycrystalline Ag film showed distinct changes in position and degradation. Although Ag is known to be an inactive material under ambient conditions, the SPR data indicated that the polycrystalline Ag film would degrade over time and thus required a dielectric protective layer.
(2010) 193–204, https://doi.org/10.1038/nmat2630. [3] D.K. Gramotnev, S.I. Bozhevolnyi, Plasmonics beyond the diffraction limit, Nat. Photonics 4 (2) (2010) 83–91, https://doi.org/10.1038/nphoton.2009.282. [4] H.A. Atwater, A. Polman, Plasmonics for improved photovoltaic devices, Nat. Mater. 9 (3) (2010) 205–213, https://doi.org/10.1038/nmat2629. [5] M. Boccas, T. Vucina, C. Araya, E. Vera, C. Ahhee, Protected-silver coatings for the 8-m Gemini telescope mirrors, Thin Solid Films 502 (1–2) (2006) 275–280, https:// doi.org/10.1016/j.tsf.2005.07.295. [6] P. Nagpal, N.C. Lindquist, S.-H. Oh, D.J. Norris, Ultrasmooth patterned metals for plasmonics and metamaterials, Science 325 (5940) (2009) 594–597, https://doi. org/10.1126/science.1174655. [7] J.H. Park, P. Ambwani, M. Manno, N.C. Lindquist, P. Nagpal, S.H. Oh, C. Leighton, D.J. Norris, Single-crystalline silver films for plasmonics, Adv. Mater. 24 (29) (2012) 3988–3992, https://doi.org/10.1002/adma.201200812. [8] H. Yu, T.P.H. Sidiropoulos, W. Liu, C. Ronning, P.K. Petrov, S.-H. Oh, S.A. Maier, P. Jin, R.F. Oulton, Influence of silver film quality on the threshold of plasmonic nanowire lasers, Adv. Opt. Mater. 5 (6) (2017) 1600856, https://doi.org/10.1002/ adom.201600856. [9] D. Yoo, K.L. Gurunatha, H.K. Choi, D.A. Mohr, C.T. Ertsgaard, R. Gordon, S.H. Oh, Low-power optical trapping of nanoparticles and proteins with resonant coaxial nanoaperture using 10 nm gap, Nano Lett. 18 (6) (2018) 3637–3642, https://doi. org/10.1021/acs.nanolett.8b00732. [10] A.M. Kern, O.J. Martin, Excitation and reemission of molecules near realistic plasmonic nanostructures, Nano Lett. 11 (2) (2011) 482–487, https://doi.org/10. 1021/nl1032588. [11] J.S. Huang, V. Callegari, P. Geisler, C. Bruning, J. Kern, J.C. Prangsma, X. Wu, T. Feichtner, J. Ziegler, P. Weinmann, M. Kamp, A. Forchel, P. Biagioni, U. Sennhauser, B. Hecht, Atomically flat single-crystalline gold nanostructures for plasmonic nanocircuitry, Nat. Commun. 1 (2010) 150, https://doi.org/10.1038/ ncomms1143. [12] E. Kazuma, K. Matsubara, K.L. Kelly, N. Sakai, T. Tatsuma, Bi- and uniaxially oriented growth and plasmon resonance properties of anisotropic Ag nanoparticles on single crystalline TiO2 surfaces, J. Phys. Chem. C 113 (2009) 4758–4762, https:// doi.org/10.1021/jp900749x. [13] H. Zhao, F. Wang, Y. Ning, B. Zhao, F. Yin, Y. Lai, J. Zheng, X. Hu, T. Fan, J. Tang, D. Zhang, K. Hu, Green “planting” nanostructured single crystal silver, Sci. Rep. 3 (2013) 1511, https://doi.org/10.1038/srep01511. [14] H.J. Lezec, A. Degiron, E. Devaux, R.A. Linke, L. Martin-Moreno, F.J. Garcia-Vidal, T.W. Ebbesen, Beaming light from a subwavelength aperture, Science 297 (5582) (2002) 820–822, https://doi.org/10.1126/science.1071895. [15] K. Yamaguchi, T. Inoue, M. Fujii, T. Ogawa, Y. Matsuzaki, T. Okamoto, M. Haraguchi, M. Fukui, Characteristics o flight intensity enhancement o fa silver nanoprism with rounded corners, J. Microsc. 229 (2008) 545–550, https://doi.org/ 10.1111/j.1365-2818.2008.01941.x. [16] T. Mori, Y. Tanaka, Y. Suzaki, K. Yamaguchi, Fabrication of silver thin film for single-crystalline nanopillar: effects of thickness and grain size, Appl. Phys. A 121 (4) (2015) 1359–1363, https://doi.org/10.1007/s00339-015-9497-1. [17] T. Mori, Y. Tanaka, Y. Suzaki, K. Yamaguchi, Advanced fabrication of single-crystalline silver nanopillar on SiO2 substrate, Appl. Phys. Lett. 108 (4) (2016) 043102, https://doi.org/10.1063/1.4940439. [18] S. Lee, J.Y. Kim, T.W. Lee, W.K. Kim, B.S. Kim, J.H. Park, J.S. Bae, Y.C. Cho, J. Kim, M.W. Oh, C.S. Hwang, S.Y. Jeong, Fabrication of high-quality single-crystal Cu thin films using radio-frequency sputtering, Sci. Rep. 4 (2014) 6230, https://doi.org/10. 1038/srep06230. [19] S. Lee, H.-Y. Park, S.J. Kim, H. Lee, I.-J. Lee, C.R. Cho, E. Lee, S.-Y. Jeong, Y.H. Lee, Inverse Stranski-Krastanov growth in single-crystalline sputtered Cu thin films for wafer-scale device applications, ACS Appl. Nano Mater. 2 (5) (2019) 3300–3306, https://doi.org/10.1021/acsanm.9b00673. [20] Y. Tsuda, H. Omoto, K. Tanaka, H. Ohsaki, The underlayer effects on the electrical resistivity of Ag thin film, Thin Solid Films 502 (1–2) (2006) 223–227, https://doi. org/10.1016/j.tsf.2005.07.279. [21] H.J. Lee, C.H. Park, S.Y. Jeong, K.J. Yee, C.R. Cho, M.H. Jung, D.J. Chadi, Hydrogen-induced ferromagnetism in ZnCoO, Appl. Phys. Lett. 88 (6) (2006) 062504, https://doi.org/10.1063/1.2171789. [22] P. Kung, C.J. Sun, A. Saxler, H. Ohsato, M. Razeghi, Crystallography of epitaxial growth of wurtzite-type thin films on sapphire substrates, J. Appl. Phys. 75 (9) (1994) 4515–4519, https://doi.org/10.1063/1.355943. [23] C.J. Sun, P. Kung, A. Saxler, H. Ohsato, K. Haritos, M. Razeghi, A crystallographic model of (00⋅1) aluminum nitride epitaxial thin film growth on (00⋅1) sapphire substrate, J. Appl. Phys. 75 (8) (1994) 3964–3967, https://doi.org/10.1063/1. 356017. [24] T. Kato, P. Kung, A. Saxler, C.-J. Sun, H. Ohsato, M. Razeghi, T. Okuda, Simultaneous growth of two differently oriented GaN epilayers on (11.0) sapphire II. A growth model of (00.1) and (10.0) GaN, J. Cryst. Growth 183 (1998) 131–139, https://doi.org/10.1016/S0022-0248(97)00364-3. [25] S.-K. Kim, S.-Y. Jeong, C.-R. Cho, Structural reconstruction of hexagonal to cubic ZnO films on Pt/Ti/SiO2/Si substrate by annealing, Appl. Phys. Lett. 82 (4) (2003) 562–564, https://doi.org/10.1063/1.1536253. [26] Y.C. Cho, S. Lee, M. Ajmal, W.-K. Kim, C.R. Cho, S.-Y. Jeong, J.H. Park, S.E. Park, S. Park, H.-K. Pak, H.C. Kim, Copper better than silver: electrical resistivity of the grain-free single-crystal copper wire, Cryst. Growth Des. 10 (6) (2010) 2780–2784, https://doi.org/10.1021/cg1003808. [27] J.Y. Kim, M.W. Oh, S. Lee, Y.C. Cho, J.H. Yoon, G.W. Lee, C.R. Cho, C.H. Park, S.Y. Jeong, Abnormal drop in electrical resistivity with impurity doping of singlecrystal Ag, Sci. Rep. 4 (2014) 5450, https://doi.org/10.1038/srep05450. [28] G. Ding, C. Clavero, D. Schweigert, M. Le, Thickness and microstructure effects in the optical and electrical properties of silver thin films, AIP Adv. 5 (11) (2015) 117234, https://doi.org/10.1063/1.4936637.
4. Conclusions To grow single-crystalline Ag film on Al2O3 substrate that had better adhesion to the substrate, we used a ZnO buffer layer based on crystallographic structure and EADM predictions. The ZnO film showed optimized growth under the 500 °C condition on the Al2O3 substrate, and was aligned epitaxially along the c-axis with grain sizes of ca. 40–50 nm. The growth of Ag film on 5 nm ZnO was optimized at 400 °C, and exhibited an ultraflat surface with an RMS value of 0.73 nm. The XRD peak intensity of the as-grown 100-nm-thick Ag film was five-fold larger than that of the Al2O3 single-crystal substrate, and its EBSD image indicated crystallographic alignment in the [1 1 1] direction. The resistivity of the single-crystalline Ag film was only about half that of the polycrystalline Ag film, and the reflectivity of the single-crystalline film was enhanced. Besides the choice of a ZnO buffer layer, the remarkable improvement in the quality of the Ag film is attributable to the modified sputtering system with a single-crystal Ag target, which enables 2- and 4-inch wafer-scale growth via the stacking of atomic layers. Thus, the results of this study provide a route to simple fabrication of materials requiring ultrahigh adhesion, anti-oxidative stability, and good optical properties for nano-patterning of Ag. CRediT authorship contribution statement Bo-Gwang Jung: Investigation, Writing - original draft, Visualization. Miyeon Cheon: Writing - review & editing, Visualization, Data curation. Su Jae Kim: Formal analysis, Data curation. Alexander Gliserin: Investigation, Methodology. Soo Hoon Chew: Investigation, Methodology. Chae Ryong Cho: Conceptualization, Validation. Seong-Gon Kim: Conceptualization, Formal analysis. Young Hee Lee: Conceptualization, Writing - review & editing. Seungchul Kim: Conceptualization, Data curation, Writing review & editing. Se-Young Jeong: Conceptualization, Data curation, Writing - review & editing, Supervision. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (No. NRF2017R1A2B3011822), (No. NRF-2017R1C1B2006137) and supported by the Institute for Basic Science (IBS-R011-D1). Appendix A. Supplementary material Supplementary data to this article can be found online at https:// doi.org/10.1016/j.apsusc.2020.145705. References [1] E. Ozbay, Plasmonics merging photonics and electronics at nanoscale dimensions, Science 311 (5758) (2006) 189–193, https://doi.org/10.1126/science.1114849. [2] J.A. Schuller, E.S. Barnard, W. Cai, Y.C. Jun, J.S. White, M.L. Brongersma, Plasmonics for extreme light concentration and manipulation, Nat. Mater. 9 (3)
7
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Full Length Article
Wafer-scale high-quality Ag thin film using a ZnO buffer layer for plasmonic applications
T
Bo-Gwang Junga, Miyeon Cheonb, Su Jae Kimb, Alexander Gliserinc, Soo Hoon Chewc, ⁎ ⁎ ⁎ Chae Ryong Chod, Seong-Gon Kime, Young Hee Leef,g, , Seungchul Kima,c, , Se-Young Jeonga,c, a
Department of Cogno-mechatronics Engineering, Pusan National University, Busan 46241, Republic of Korea Crystal Bank Research Institute, Pusan National University, Busan 46241, Republic of Korea c Department of Optics and Mechatronics Engineering, Pusan National University, Busan 46241, Republic of Korea d Department of Nanoenergy Engineering, Pusan National University, Busan 46241, Republic of Korea e Department of Physics and Astronomy, Mississippi State University, Mississippi State, MS 39762, USA f Center for Integrated Nanostructure Physics, Institute for Basic Science, Sungkyunkwan University Suwon 16419, Republic of Korea g Department of Energy Science, Sungkyunkwan University, Suwon 16419, Republic of Korea b
A R T I C LE I N FO
A B S T R A C T
Keywords: Ag thin film EADM Reflectivity Surface plasmon ZnO buffer layer
Realizing laterally continuous, ultraflat silver (Ag) single-crystal films is a significant technological challenge. Ag thin film grown on various hetero-substrates has been used in numerous applications, due to its superior electrical and optical properties. To exploit these properties without degradation and apply these films to highprecision patterning, surface plasmonics, and so on, a high-quality thin film having an ultraflat surface and few grain boundaries is needed. A zinc oxide (ZnO) buffer layer can be used to facilitate the growth of a singlecrystalline Ag thin film on a sapphire (Al2O3) substrate. ZnO films deposited on Al2O3 substrates have grain boundaries; however, Ag films grown on ZnO are nearly grain-free and close to single-crystalline quality. This can be explained by the exceptionally small extended atomic distance mismatch, of ~0.08%, between Ag and Al2O3, in which the ninth Ag atom and the eighth Al atom are matched in terms of coherence and periodicity. A modified radio frequency sputtering system with a single-crystal Ag target enabled wafer-scale growth of ultraflat, grain-free Ag films. The proposed approach using a ZnO buffer provides a new method for fabricating Ag films with high adhesion, anti-oxidative stability, and superior optical properties, and allows for easy nanopatterning.
1. Introduction Noble metal films, such as gold (Au) and silver (Ag) films, have attracted considerable interest in optical nano-circuitry, next-generation photovoltaic devices, and surface plasmonic resonance [1–4]. Agcoated mirrors are ideal for observing a wide range of light wavelengths, having high reflectivity and low emissivity that is especially useful in infrared applications for astronomy [5]. Because electromagnetic coupling between light and free electrons at metal surfaces is important in surface plasmons, surface roughness and other inhomogeneities of the metal that limit surface-plasmon propagation are a major barrier to efficient plasmonic devices [6–9]. In a study comparing the optical response of two plasmonic nanoantennae, distinct differences between two structures, where one had an idealized geometry and the other a realistic geometry, resulted in
approximately 1.4-fold differences in the near-field properties, polarization, and spectral behavior [10]. Deep subwavelength integration of high-definition plasmonic nanostructures, which is of key importance in the development of future optical nanocircuitry, depends on commonly used thermally evaporated Au layers. Evaporated Au layers tend to have structural imperfections that appear in individual circuit elements, resulting in a drastic reduction in the yield of integrated nanocircuits. Chemically grown single-crystalline Au flakes have been shown to provide an ideal basis for focused ion beam (FIB) milling and other topdown nanofabrication techniques, to produce high-definition, ultrasmooth Au nanostructures with superior optical properties [11]. Although the fabrication of single-crystalline metallic nanostructures using chemical processing methods has been reported [12,13], a physical technique, such as evaporation or sputtering, is preferred, because chemical processing methods make it difficult to control uniformity due
⁎ Corresponding authors at: Center for Integrated Nanostructure Physics, Institute for Basic Science, Sungkyunkwan University Suwon 16419, Republic of Korea (Y.H. Lee). Department of Cogno-mechatronics Engineering, Pusan National University, Busan 46241, Republic of Korea (S. Kim and S.-Y. Jeong). E-mail addresses: [email protected] (Y.H. Lee), [email protected] (S. Kim), [email protected] (S.-Y. Jeong).
https://doi.org/10.1016/j.apsusc.2020.145705 Received 20 December 2019; Received in revised form 30 January 2020; Accepted 8 February 2020 Available online 08 February 2020 0169-4332/ © 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Applied Surface Science 512 (2020) 145705
B.-G. Jung, et al.
white plate (barium sulfate). The incident angle of the beam was set at 5°.
to the metallic nanoclusters that form and subsequently complicate the synthesis process [14,15]. Extremely flat, single-crystalline Ag films with a root mean square (RMS) roughness of 0.7 nm and grain size of several hundred nanometers have been obtained using a standard direct current magnetron sputtering system; the films possessed several key properties necessary for plasmonics applications [7]. A single-crystalline Ag film grown on a SiO2 substrate was used to obtain a Ag nanopillar structure by FIB milling; [16,17] observation of the crystal orientation of the Ag films using an electron backscatter diffraction (EBSD) method confirmed grain growth, achieved by adjusting the film deposition conditions. However, the grain size under the optimum conditions was no more than ~200 nm. Thus, growth of grain-boundary-free single-crystalline metal films and replacement of Au film with copper or Ag single-crystalline film remains challenging. Sputtering is expected to provide an innovative solution to the aforementioned problems and enable waferscale production. In our previous studies, we introduced a method to obtain highquality wafer-scale grain-boundary-free single-crystalline copper thin films on a sapphire (Al2O3) substrate using a sputtering method [18,19]. However, Ag films grown by the same method showed unexpectedly poor adhesion to the Al2O3 substrate due to the large atomic distance mismatch (ADM), which led to film delamination from the substrate. In this letter, we introduce the growth of ultraflat, wafer-scale single-crystalline Ag thin films by incorporating a zinc oxide (ZnO) buffer layer on the Al2O3 substrate. We describe how the Ag atoms overcome the large ADM, with the substrate using the ZnO buffer layer to relieve the large strain induced by heteroepitaxy, thus resulting in single-crystalline film with an ultraflat surface and high crystallinity.
3. Results and discussion Ag has a large ADM of 4.9% or 5.1% with respect to the Al2O3 substrate, varying depending on the growth direction. Despite efforts to optimize film growth conditions, Ag film on the Al2O3 (0 0 6) substrate is still of poor quality, and separates easily from the substrate. The quality of Ag films grown on a Al2O3 (0 0 6) substrate with previously reported conventional methods was not sufficient for plasmonic applications. The typical crystalline properties of our Ag film grown on Al2O3 are presented in Fig. S1 (Supporting Information). The EBSD orientation map (Fig. S1a, Supporting Information), its inverse pole figure (IPF) (Fig. S1b, Supporting Information) [3], pole figure (PF) with respect to the normal direction, i.e., ∥ [1 1 1] of Ag (Fig. S1c, Supporting Information), and X-ray diffraction (XRD) pattern (Fig. S1f, Supporting Information) show the polycrystalline nature of the Ag thin film on Al2O3, despite the use of a modified sputtering method. Every grain in the polycrystalline Ag was aligned in an arbitrary direction, and the grain size was, on average, several tens of nanometers. Scanning electron microscopy (SEM) images obtained under low (Fig. S1d, Supporting Information) and high magnification (Fig. S1e, Supporting Information) confirmed the EBSD results, i.e., that the grains of the Ag film are deposited in arbitrary directions with a grain size of several tens of nanometers. The surface roughness obtained from atomic force microscopy (AFM) measurements (Fig. S1g, Supporting Information) indicated an RMS value of 5.9 nm, which increased over time due to further oxidation in a normal laboratory environment. Based on these results, we concluded that the Al2O3 (0 0 6) substrate was not satisfactory for Ag film growth, and we proceeded to identify a reasonable buffer layer to reduce lattice mismatch and release the strain induced by the large ADM. Given that we could not improve the crystallinity of the Ag film using a Si substrate, we chose Al2O3 as the substrate material. ZnO was determined to be the best candidate as a buffer layer for the Al2O3 substrate, as it shows a good structural match with Al2O3 and is chemically stable. ZnO film has been grown on Al2O3 (0 0 6) substrates for lightemitting diode (LED) and dilute magnetic semiconductor applications, and is well known to be compatible with Al2O3 (0 0 6) substrates [20,21]. The schematic structure of an Ag thin film, on a ZnO buffer layer on a Al2O3 (0 0 6) substrate, and an actual photograph of a 50-nmthick Ag thin film grown on a ZnO buffer layer are shown in Fig. 1a. XRD images were taken of the areas marked by the colored squares in Fig. 1a; the color of the curves in Fig. 1b corresponds to the area marked by the same color in Fig. 1a. The similar XRD results among the different areas prove the wafer-scale homogeneity of the Ag thin film on the ZnO buffer layer. The peaks at 34.4°, 38.2°, and 41.7° correspond to those of ZnO (0 0 2), Ag(1 1 1), and Al2O3 (0 0 6), respectively. An AFM image of the Ag film (Fig. 1c) shows a very flat surface in the vertical scale of a full height of 20 nm; for the 10 µm × 10 µm area; the RMS surface roughness value was 0.73 nm, slightly larger than the RMS roughness (~0.26 nm) of the Al2O3 substrate. The right upper inset shows an AFM image of a 2 µm × 2 µm area with a higher contrast in the vertical scale of 5 nm, which was acquired to better resolve the surface roughness. The Al2O3 substrate-ZnO buffer layer-Ag film interfaces were characterized using cross-sectional high-resolution transmission electron microscopy (HR-TEM). The cross-sectional HRTEM images with different resolutions reveal epitaxial growth of the Ag film (Fig. 1d–f). The fast Fourier transform (FFT) patterns of the areas in Fig. 1f marked with rectangular boxes are shown in Fig. 1g–i. The orientation of the Al2O3 substrate parallel to the electron beam direction is 〈110〉, whereas the orientations of the ZnO buffer layer and the Ag film are 〈21¯0〉 and 〈110〉, respectively. The (0 0 6) planes of Al2O3, ZnO
2. Materials and methods Ag thin films were fabricated using a modified radio frequency sputtering method. The single-crystal Ag target used in the sputtering process was prepared as 2-inch-diameter disks derived from singlecrystal Ag ingots grown by the Czochralski method, supported by the Crystal Bank of Pusan National University (Busan, Korea). Special care was taken to reduce the instability of the sputtering system caused by the electrical power supply and mechanical vibrations. Ag thin films were grown on 5 nm thick ZnO thin film sputtered on a 2-inch Al2O3 (0001) substrate under various conditions. The base pressure was ~10−6 Torr, and the working pressure, of 10−3, was adjusted by highpurity Ar gas (99.999%). The growth rate was 1 Å/s. The θ-2θ XRD measurements were performed using an Empyrean Series 2 instrument (PANalytical B.V., Almelo, The Netherlands) equipped with a Cu-Kα source (40 kV, 30 mA). Data were collected within the range of 20° < 2θ < 80°, with a step size of 0.0167°. AFM measurements were carried out using an XE-100 instrument (Park Systems, Inc., Suwon, Korea). AFM images of the Ag thin films (10 × 10 µm2) were obtained in non-contact mode (cantilever: PPPNCHR). The SEM and EBSD images were obtained using a SUPRA40 VP system (Carl Zeiss AG, Oberkochen, Germany). The PF represents the spatial orientation of grains and the orientation distribution of crystallographic lattice planes in stereographic projection. The IPF shows the pole that is parallel to a given thin film direction. The IPF represents only a unit triangle, as opposed to a full circle, due to crystal symmetry. The IPF map combines the locally detected orientation with the crystallographic data. The color coding is in accordance with the reduced size of the IPF. The electrical resistivity of the Ag thin film was measured using the van der Pauw method (HMS-3000; ECOPIA, Anyang, Korea) and the diffuse reflectance was measured using an ultraviolet-visible spectrometer (V-770; Jasco Inc., Easton, MD, USA) equipped with a 60-mm Φ integrating sphere. The reflectivity was measured over the range 180–1000 nm. The baseline correction procedure used a reference 2
Applied Surface Science 512 (2020) 145705
B.-G. Jung, et al.
Fig. 1. Structural details of a single-crystalline silver (Ag) thin film (thickness: 50 nm) grown on a 5 nm zinc oxide (ZnO) buffer film. (a) Schematic diagram and photographic images of an Ag film on ZnO/sapphire (Al2O3). (b) X-ray diffraction (XRD) patterns corresponding to the areas marked in (a). (c) Atomic force microscopy (AFM) images of a single-crystalline Ag thin film on the ZnO buffer layer with the vertical scale of the full height of 20 nm (inset with 5 nm). (d–f) Crosssectional high-resolution transmission electron microscopy (HR-TEM) images, under different magnifications, of a single-crystalline Ag thin film on a ZnO buffer layer. (g–i) Fast Fourier transform (FFT) patterns of Al2O3, ZnO, and Ag, corresponding to the areas in the HR-TEM image (f) marked by rectangular boxes.
exhibited a dark gray color. The XRD and SEM results of the Ag thin films grown on the ZnO buffer layer deposited at 500 °C are shown in Fig. 2c and d, respectively. The thickness of Ag films was adjusted to ~100 nm during the optimization process. In the XRD data presented at the linear scale (Fig. 2c), only three peaks appeared, corresponding to the ZnO (0 0 2) buffer layer, Ag (1 1 1) film, and Al2O3 (0 0 6) substrate at 34.4°, 38.2°, and 41.7°, respectively. The intensity of XRD for each sample was normalized with respect to the Al2O3 (0 0 6) intensity, to estimate the crystallinity of the Ag (1 1 1) peak. The intensity of the Ag (1 1 1) peak for all samples was at least three-fold higher than the peak intensity of the Al2O3(0 0 6) substrate itself, whereas the Ag (1 1 1) peak of the polycrystalline Ag film grown directly on the Al2O3(0 0 6) substrate (Fig. S1f, Supporting Information) was almost hidden by the Al2O3(0 0 6) substrate peak, and was not observed at all at the linear scale. Considering the highest peak intensity and smallest full width half maximum of the Ag (1 1 1) (0.18°) peak, the optimum condition for Ag film growth on ZnO was taken to be 400 °C in this study. Fig. 2d shows SEM images of the samples grown at 200 °C, 300 °C, 400 °C, and 500 °C. Considering that the ZnO buffer layer was aligned with the [1 0 0] direction, but was still polycrystalline, the smooth surfaces of all four samples observed in low-magnification SEM images (1-μm scale; insets in Fig. 2d) were surprising. A clear grain boundary could not be identified for the sample grown at 400 °C, even under high magnification. This suggests that the Ag film was nearly single-crystalline. We examined the RMS values of Ag films as a function of growth temperature (Fig. S2b, Supporting Information), where the RMS value was 1.50 nm for the sample grown at 200 °C and 0.73 nm for that grown at 400 °C. The right panel of the AFM images indicates the line profiles across the AFM surface images, and the RMS surface roughness values calculated for a 10 µm × 10 µm area of the thin film. These images further confirmed the single-crystal-like growth structure on the polycrystalline substrate, for which heteroepitaxy is very unusual. To understand the growth mechanism, it is necessary to investigate the crystallographic structure of the grown Ag film in detail. To determine the crystal orientation in the plane, and to understand the grain growth process, we examined coincident site lattice (CSL) boundary
(0 0 1), and Ag (1 1 0) are normal with respect to the interfaces; the respective normal values for these directions, 〈21¯0〉 of Al2O3, 〈010〉 of ZnO and 〈11¯0〉 of Ag, are also parallel to each other. Thus, HR-TEM analysis supports the observed growth in the 〈111〉 direction of the Ag thin film on a Al2O3 substrate. We attempted to grow the ZnO buffer film on Al2O3 substrates under various growth temperatures, ranging from 200 °C to 500 °C, to determine the optimum growth conditions based on XRD and SEM measurements. The films were adjusted to have the same thickness, of ~100 nm. The XRD results for the ZnO buffer films are presented on a linear scale; the intensities of ZnO peaks were normalized with respect to the Al2O3 (0 0 6) intensity in Fig. 2a. All samples showed clear ZnO (0 0 2) and Al2O3 (0 0 6) peaks at 34.4° and 41.7°, respectively, which indicates that the ZnO films were well aligned in the [0 0 1] direction along the Al2O3 (0 0 6) plane. The two samples grown at 400 °C and 500 °C showed a single phase and high crystallinity. However, SEM revealed larger grains in the sample grown at 500 °C (Fig. 2b). Lowmagnification SEM images of the ZnO (insets in Fig. 2b) confirmed that the grown ZnO films were of good quality and uniform, regardless of the growth temperature over the range 200–500 °C. High-magnification SEM images clearly revealed grain growth (20–50 nm in size), where grain size increased with growth temperature up to 500 °C. At temperatures higher than 500 °C, the film quality of ZnO deteriorated (not shown here). Based on these results, we limited the growth temperature of the ZnO film to below 500 °C. Notably, the ZnO film grown at 500 °C showed the best crystallinity and largest grain size among the film samples; additionally, the ZnO films grown at different temperatures showed no clear difference in RMS roughness (Fig. S2a, Supporting Information). Thus, 500 °C was chosen as the optimal growth temperature of the ZnO buffer layer with a thickness of 5 nm for Ag film growth; all other growth conditions, such as the initial pressure, working pressure, Ar flow, and film thickness were held constant during Ag film growth. Once again, we varied the growth temperature of the Ag film over the range of 200–500 °C to identify the optimal Ag film growth conditions. Notably, Ag film grown above 600 °C showed oxidization and 3
Applied Surface Science 512 (2020) 145705
B.-G. Jung, et al.
Fig. 2. Optimization of the growth temperature of a ZnO buffer film and Ag film. (a, b) XRD (a) and scanning electron microscopy (SEM) images (b) of ZnO films. ZnO buffer films deposited on Al2O3 (0 0 6) substrates; the optimum condition was 500 °C. (c, d) XRD (c) and SEM images (d) of Ag thin films deposited on a 5 nm ZnO buffer layer. The single-crystalline Ag thin films were grown on a Al2O3 substrate using a ZnO buffer layer under the optimum condition of 400 °C.
line maps of Ag thin films using EBSD. Crystallographic characterization of Ag thin films was performed. The CSL boundary lines in the EBSD maps of Ag thin films, grown on a Al2O3 substrate using a ZnO buffer layer, showed the changes in film crystallinity with growth temperature (Fig. 3a); the blue color corresponds to the crystal orientation along the 〈111〉 direction, and the CSL boundary lines denote the boundaries between areas with different orientations. The grain size was estimated based on the CSL boundary lines; the entire measured area of all samples was aligned, or nearly aligned, with the 〈111〉 direction, in contrast to the polycrystalline sample (Fig. S1a, Supporting Information). The IPFs corresponding to the EBSD orientation maps shown in Fig. 3a show the degree to which each grain is aligned with the (1 1 1) plane (Fig. 3b); the dispersion of the spots in the area near (1 1 1) was caused by the tilting of two twins with different stacking orders.[19] The CSL boundary lines in the EBSD maps showed that the sample grown at 200 °C had relatively smaller grains. The grains of samples grown at 300 °C and 500 °C were larger, but showed a slight difference in degree of tilt from the (1 1 1) plane; thus, the spots in IPF show a wide dispersion. The six-fold symmetry in the EBSD {1 0 0} PF with respect to the normal direction, i.e., ∥ [1 1 1] (Fig. 3c), arises from the two grains rotated by 60° with an ABCABC⋯ and ACBACB⋯ stacking order [19]. Although the spots in PF (Fig. 3c, insets) were located within a small area, the spots in the IPF were widely dispersed or separated into two clusters, as each of the three originates from two grain types of different stacking orders. The IPF under the 400 °C condition was better aligned in the (1 1 1) direction; two grain types (same color on the EBSD map) were well aligned in the (1 1 1) plane. The CSL boundary lines in the EBSD maps are represented in red, and the rotation angle between the two grains ranged between 59°and 60°. Parts of the red CSL boundary lines were larger than 99.5% for the sample grown at 400 °C (see the rotation angles below the corresponding EBSD
map in Fig. 3), which implies that two grain types were typically rotated by nearly 60° with respect to the other. Ideally, the {1 0 0} PF of the cubic structure with respect to the normal direction, i.e., ∥ [1 1 1], should have three-fold symmetry; however, the six-fold symmetry of the φ-scan XRD pattern of the Ag (1 1 1) plane (Fig. 3d) implies 60° rotation of two types of grains in the thin film. The peaks marked by ▲ and ▽ on the XRD φ-scan represent the two grain types; the equivalent markers in the EBSD map obtained under the 400 °C condition indicate the cause of the six-fold symmetry in the {1 0 0} PF in Fig. 3c, and the XRD peaks in Fig. 3d, respectively. The formation of the two kinds of grains rotated by 60° was attributed to the two stacking orders, ABC… and ACB…, in the face-centered cubic (1 1 1) plane (Fig. 3e). These two grains are connected by a hidden incoherent twin boundary (HITB) that does not permit vacant atoms; thus, the boundary assists overlayer growth without defects. The mechanism of HITB formation, and the growth of the grains to form a single grain, were first discussed in a study of copper single-crystal thin films; [19] interestingly, the Ag thin film is grown on the ZnO buffer layer according to the same mechanism by which copper thin film is grown on Al2O3. Ag film growth on the ZnO buffer layer of an Al2O3 substrate involved two types of heteroepitaxy. The ZnO film worked well as a buffer; however, it retained a polycrystalline form and, thus was not single-crystal in nature to the same extent as the Al2O3 substrate. Furthermore, the large ADM between Ag and ZnO makes it difficult to explain how Ag thin film of single-crystalline quality can grow on the polycrystalline ZnO buffer layer. Thus, we used an EADM model to determine whether there was extended periodicity of the lattice match between the two materials. The structural relations among the substrate, buffer, and final growth material are illustrated by the superposition of atomic layers in Fig. 4a. According to the Moire pattern observed with lattice superposition, overlayer growth can be predicted. 4
Applied Surface Science 512 (2020) 145705
B.-G. Jung, et al.
Fig. 3. Crystallographic orientations of Ag films grown on a ZnO buffer film under various temperatures: (a) coincident site lattice (CSL) boundary line maps of electron backscatter diffraction (EBSD) measurements of Ag thin films. All scale bars are 1 μm; (b, c) inverse pole figure (IPF) corresponding to the EBSD orientation maps of (a) (b) and EBSD [4] pole figure (PF) with respect to the normal direction, i.e., ∥ Ag [1 1 1] (c). (d, e) φ-scan XRD pattern of the Ag (1 1 1) plane (d) and an atomic model showing two possible stacking configurations on the hexagonal basal plane, in which a grain is rotated by 60˚ (e).
the epilayer [23]. The lattice mismatch calculated between ZnO and Al2O3 was ~18%, as shown in Fig. 4b. The overlapping crystal structures of ZnO and Al2O3 indicate that these two materials have extended periodicity; however, the respective EADMs were 0.51% and 1.5%, which are still relatively large. Based on this result, it is inevitable that the ZnO buffer layer grown on Al2O3 has grain boundaries, even under optimized conditions (Fig. 3b). On the other hand, Ag shows remarkable similarity to ZnO; while the ADM was ~11% for both materials, the EADM was ~0.08%. Because nine times the Ag-Ag distance of Ag is approximately equal to eight times the ZnZn distance of ZnO, I and I′ are 9 and 8, respectively, and the larger of the period lengths in two directions is about 5.0 nm, which is much smaller than that between Ag and Al2O3. The small EADM, I and I′ values promote ultraflat growth of Ag film on ZnO with single-crystalline quality. All results of ADM and EADM are summarized in Table 1. The electrical conductivity and optical reflectivity are representative physical quantities sensitive to grain boundary formation. The scattering of electrons at grain boundaries and rough surfaces is the main cause of resistivity and deterioration of electric properties. The reduction in resistivity achieved by eliminating grain boundaries amounts to about 13% in bulk copper and Ag substrates [26,27]. The
Generally, oxides such as Al2O3 are ionic crystals, in which small-size cations, such as A13+, with larger valences easily attract and retain oxygen ions. The surface of Al2O3 consists only of oxygen ions [22]. Thus, the lattice mismatch can be calculated with respect to the change in Ag-O atomic distances between Ag and Al2O3. The O-O distance of ¯ 10] Al2O3 in the [1 0 0] direction was 2.753 Å, and that of Ag in the [1 direction was 2.889 Å. The degree of ADM was 4.9%, with 20:21 being the smallest integral ratio (2.889:2.753). Twenty times the Ag-Ag distance of Ag is nearly equal to twenty-one times that of the O-O distance of Al2O3, as shown in Fig. 4a. The EADM for large lattice mismatch epitaxial growth can be calculated as follows: EADM = Id − I 'd' , where I and I′ are integers and d and d′ are atomic I 'd' distances of the epilayer and substrate, respectively. I and I′ are determined as follows: d:d′ ~ I:I′, where I:I′ is the smallest integral ratio for d:d′. The difference between I and I′ introduces a periodic edge-type dislocation [23–25]. Although the EADM for Ag and Al2O3 is small (0.08%), I and I′ are 20 and 21, respectively. This period length in Ag in the [1¯10 ] vertical direction is about 5.78 nm, and that in the [112¯ ] horizontal direction is about 10.01 nm, which is sufficiently long for other defects to potentially appear. The lowest mismatch limit that can be applied to EADM is not fixed, and depends on the deformation potential of the epilayer, as well as on the nature and number of defects in
Fig. 4. Extended atomic distance mismatch between the film and substrate lattice: superposition of (a) Ag in Ag (1 1 1) on O in Al2O3 (0 0 1), (b) Zn in ZnO (0 0 2) on O in Al2O3 (0 0 1), and (c) Ag in Ag (1 1 1) on O in ZnO (0 0 2). 5
Applied Surface Science 512 (2020) 145705
B.-G. Jung, et al.
by the absolute decrease in electron scattering seen at the grain boundaries and on rough surfaces. The reflectivity of the single-crystalline 50-nm-thick Ag thin film was compared to that of a commercial polycrystalline 150-nm-thick Ag thin film (G-mek, Inc., Gyeonggi, Korea). The single-crystalline Ag thin film showed clearly superior reflectivity over the measured wavelength range. To determine the chemical stability of the single-crystalline Ag thin film, its reflectivity was measured again 6 months later and similar results were obtained [Fig. 5(b) and Supplementary Fig. S3]. The fact that the reflectivity of our Ag film was maintained even after several months indicates the chemical stability of the film; notably, a protective layer for commercial Ag mirrors is mandatory. Surface plasmon resonance (SPR) curves were obtained for singleand polycrystalline Ag films using a coherent 632.8-nm light source (Fig. 5c). In the single-crystalline Ag film, a deeper and sharper resonance curve was obtained compared to that for the polycrystalline Ag film, with the same film thickness. The position and depth of the dip in SPR well-matched that obtained with numerical simulations of a singlecrystalline Ag film (Supplementary Fig. S4). This implies that the singlecrystalline film should have a higher real permittivity and lower imaginary permittivity. The width of the plasmonic dip depends on the imaginary part of the dielectric constant of the metal. Such an SPR angle mismatch between single- and polycrystalline films may originate from the ZnO interlayer film below the single-crystalline Ag film. However, numerical simulations showed that the ZnO buffer layer did not affect the SPR curve (Supplementary Fig. S4). The chemical stability of the single- and polycrystalline Ag films was evaluated based on the SPR curve, after exposing the samples to ambient conditions for 2 months (Fig. 5d and e). The shift in SPR curve of the single-crystalline
Table 1 Atomic distance mismatch (ADM) and extended atomic distance mismatch (EADM). Material
Atomic distance (Å)
ADM
I
I′
EADM
Ag(1 1 1)/Al2O3(0 0 6)
D1 = 2.889 D2*=2.753 D′1 = 5.004 D′2 = 4.761 D3 = 3.248 D2 = 2.753 D′3 = 5.625 D′2 = 4.761 D1 = 2.889 D3 = 3.248 D′1 = 5.004 D′3 = 5.625
4.9% 5.1% 18.0% 18.1% 11.1% 11.0%
20 20 6 5 9 9
21 21 7 6 8 8
0.08% 0.1% 0.51% 1.5% 0.07% 0.08%
ZnO(0 0 2)/Al2O3(0 0 6) Ag(1 1 1)/ZnO(0 0 2)
D1, D2, and D3 are the vertical atomic distances of Ag-Ag, O-O in Al2O3, Zn(O)Zn(O) respectively. D′1, D′2, and D′3 are the horizontal atomic distances of AgAg, O-O in Al2O3, Zn(O)-Zn(O), respectively. * The mean distance of vertical direction oxygen atoms at Al2O3 was calculated as (2.868 × 2 + 2.524)/ 3 = 2.753 (Å).
minimum resistivity that can be achieved in a copper single-crystal film was reported as 1.61 μΩ·cm, which is 7% lower than the resistivity of Cu bulk according to the International Annealed Copper Standard (IACS: 1.7241 μΩ·cm), and 6% higher than that of single-crystal Cu bulk (1.52 μΩ·cm at 293 K) [18]. For our Ag thin films deposited on ZnO, no clear grain boundaries were observed. The film surface was ultraflat, with an RMS roughness nearly a quarter of that of polycrystalline Ag films (Fig. S2a, Supporting Information). In this study, the measured resistivity of the single-crystalline Ag thin films (thickness: 50 nm) was < 3 µΩ·cm; the resistivity of the samples grown under 400 °C and 500 °C was ~1.87 µΩ·cm, which is close to that of Ag bulk (1.6 µΩ·cm); the typical resistivity of polycrystalline Ag is 2.61 µΩ·cm [28] (Fig. 5a). These results suggest that the mean free path of electrons was increased
Fig. 5. Physical properties of Ag single-crystalline film grown using a ZnO buffer layer. (a) Resistivity of the Ag films as a function of growth temperature. (b) Reflectivity of the Ag thin film as a function of wavelength compared to a commercial Ag thin film. (c) Surface plasmon resonance (SPR) reflectivity of the singlecrystalline Ag film with a ZnO buffer layer of 5 nm thickness as a function of incident angle just after deposition compared to a polycrystalline Ag thin film without a ZnO buffer layer. (d, e) SPR reflectivity of a polycrystalline Ag thin film without a ZnO buffer layer (d) and a single-crystalline Ag film after 2 months (e). 6
Applied Surface Science 512 (2020) 145705
B.-G. Jung, et al.
Ag film was negligible; however, the SPR curve of the polycrystalline Ag film showed distinct changes in position and degradation. Although Ag is known to be an inactive material under ambient conditions, the SPR data indicated that the polycrystalline Ag film would degrade over time and thus required a dielectric protective layer.
(2010) 193–204, https://doi.org/10.1038/nmat2630. [3] D.K. Gramotnev, S.I. Bozhevolnyi, Plasmonics beyond the diffraction limit, Nat. Photonics 4 (2) (2010) 83–91, https://doi.org/10.1038/nphoton.2009.282. [4] H.A. Atwater, A. Polman, Plasmonics for improved photovoltaic devices, Nat. Mater. 9 (3) (2010) 205–213, https://doi.org/10.1038/nmat2629. [5] M. Boccas, T. Vucina, C. Araya, E. Vera, C. Ahhee, Protected-silver coatings for the 8-m Gemini telescope mirrors, Thin Solid Films 502 (1–2) (2006) 275–280, https:// doi.org/10.1016/j.tsf.2005.07.295. [6] P. Nagpal, N.C. Lindquist, S.-H. Oh, D.J. Norris, Ultrasmooth patterned metals for plasmonics and metamaterials, Science 325 (5940) (2009) 594–597, https://doi. org/10.1126/science.1174655. [7] J.H. Park, P. Ambwani, M. Manno, N.C. Lindquist, P. Nagpal, S.H. Oh, C. Leighton, D.J. Norris, Single-crystalline silver films for plasmonics, Adv. Mater. 24 (29) (2012) 3988–3992, https://doi.org/10.1002/adma.201200812. [8] H. Yu, T.P.H. Sidiropoulos, W. Liu, C. Ronning, P.K. Petrov, S.-H. Oh, S.A. Maier, P. Jin, R.F. Oulton, Influence of silver film quality on the threshold of plasmonic nanowire lasers, Adv. Opt. Mater. 5 (6) (2017) 1600856, https://doi.org/10.1002/ adom.201600856. [9] D. Yoo, K.L. Gurunatha, H.K. Choi, D.A. Mohr, C.T. Ertsgaard, R. Gordon, S.H. Oh, Low-power optical trapping of nanoparticles and proteins with resonant coaxial nanoaperture using 10 nm gap, Nano Lett. 18 (6) (2018) 3637–3642, https://doi. org/10.1021/acs.nanolett.8b00732. [10] A.M. Kern, O.J. Martin, Excitation and reemission of molecules near realistic plasmonic nanostructures, Nano Lett. 11 (2) (2011) 482–487, https://doi.org/10. 1021/nl1032588. [11] J.S. Huang, V. Callegari, P. Geisler, C. Bruning, J. Kern, J.C. Prangsma, X. Wu, T. Feichtner, J. Ziegler, P. Weinmann, M. Kamp, A. Forchel, P. Biagioni, U. Sennhauser, B. Hecht, Atomically flat single-crystalline gold nanostructures for plasmonic nanocircuitry, Nat. Commun. 1 (2010) 150, https://doi.org/10.1038/ ncomms1143. [12] E. Kazuma, K. Matsubara, K.L. Kelly, N. Sakai, T. Tatsuma, Bi- and uniaxially oriented growth and plasmon resonance properties of anisotropic Ag nanoparticles on single crystalline TiO2 surfaces, J. Phys. Chem. C 113 (2009) 4758–4762, https:// doi.org/10.1021/jp900749x. [13] H. Zhao, F. Wang, Y. Ning, B. Zhao, F. Yin, Y. Lai, J. Zheng, X. Hu, T. Fan, J. Tang, D. Zhang, K. Hu, Green “planting” nanostructured single crystal silver, Sci. Rep. 3 (2013) 1511, https://doi.org/10.1038/srep01511. [14] H.J. Lezec, A. Degiron, E. Devaux, R.A. Linke, L. Martin-Moreno, F.J. Garcia-Vidal, T.W. Ebbesen, Beaming light from a subwavelength aperture, Science 297 (5582) (2002) 820–822, https://doi.org/10.1126/science.1071895. [15] K. Yamaguchi, T. Inoue, M. Fujii, T. Ogawa, Y. Matsuzaki, T. Okamoto, M. Haraguchi, M. Fukui, Characteristics o flight intensity enhancement o fa silver nanoprism with rounded corners, J. Microsc. 229 (2008) 545–550, https://doi.org/ 10.1111/j.1365-2818.2008.01941.x. [16] T. Mori, Y. Tanaka, Y. Suzaki, K. Yamaguchi, Fabrication of silver thin film for single-crystalline nanopillar: effects of thickness and grain size, Appl. Phys. A 121 (4) (2015) 1359–1363, https://doi.org/10.1007/s00339-015-9497-1. [17] T. Mori, Y. Tanaka, Y. Suzaki, K. Yamaguchi, Advanced fabrication of single-crystalline silver nanopillar on SiO2 substrate, Appl. Phys. Lett. 108 (4) (2016) 043102, https://doi.org/10.1063/1.4940439. [18] S. Lee, J.Y. Kim, T.W. Lee, W.K. Kim, B.S. Kim, J.H. Park, J.S. Bae, Y.C. Cho, J. Kim, M.W. Oh, C.S. Hwang, S.Y. Jeong, Fabrication of high-quality single-crystal Cu thin films using radio-frequency sputtering, Sci. Rep. 4 (2014) 6230, https://doi.org/10. 1038/srep06230. [19] S. Lee, H.-Y. Park, S.J. Kim, H. Lee, I.-J. Lee, C.R. Cho, E. Lee, S.-Y. Jeong, Y.H. Lee, Inverse Stranski-Krastanov growth in single-crystalline sputtered Cu thin films for wafer-scale device applications, ACS Appl. Nano Mater. 2 (5) (2019) 3300–3306, https://doi.org/10.1021/acsanm.9b00673. [20] Y. Tsuda, H. Omoto, K. Tanaka, H. Ohsaki, The underlayer effects on the electrical resistivity of Ag thin film, Thin Solid Films 502 (1–2) (2006) 223–227, https://doi. org/10.1016/j.tsf.2005.07.279. [21] H.J. Lee, C.H. Park, S.Y. Jeong, K.J. Yee, C.R. Cho, M.H. Jung, D.J. Chadi, Hydrogen-induced ferromagnetism in ZnCoO, Appl. Phys. Lett. 88 (6) (2006) 062504, https://doi.org/10.1063/1.2171789. [22] P. Kung, C.J. Sun, A. Saxler, H. Ohsato, M. Razeghi, Crystallography of epitaxial growth of wurtzite-type thin films on sapphire substrates, J. Appl. Phys. 75 (9) (1994) 4515–4519, https://doi.org/10.1063/1.355943. [23] C.J. Sun, P. Kung, A. Saxler, H. Ohsato, K. Haritos, M. Razeghi, A crystallographic model of (00⋅1) aluminum nitride epitaxial thin film growth on (00⋅1) sapphire substrate, J. Appl. Phys. 75 (8) (1994) 3964–3967, https://doi.org/10.1063/1. 356017. [24] T. Kato, P. Kung, A. Saxler, C.-J. Sun, H. Ohsato, M. Razeghi, T. Okuda, Simultaneous growth of two differently oriented GaN epilayers on (11.0) sapphire II. A growth model of (00.1) and (10.0) GaN, J. Cryst. Growth 183 (1998) 131–139, https://doi.org/10.1016/S0022-0248(97)00364-3. [25] S.-K. Kim, S.-Y. Jeong, C.-R. Cho, Structural reconstruction of hexagonal to cubic ZnO films on Pt/Ti/SiO2/Si substrate by annealing, Appl. Phys. Lett. 82 (4) (2003) 562–564, https://doi.org/10.1063/1.1536253. [26] Y.C. Cho, S. Lee, M. Ajmal, W.-K. Kim, C.R. Cho, S.-Y. Jeong, J.H. Park, S.E. Park, S. Park, H.-K. Pak, H.C. Kim, Copper better than silver: electrical resistivity of the grain-free single-crystal copper wire, Cryst. Growth Des. 10 (6) (2010) 2780–2784, https://doi.org/10.1021/cg1003808. [27] J.Y. Kim, M.W. Oh, S. Lee, Y.C. Cho, J.H. Yoon, G.W. Lee, C.R. Cho, C.H. Park, S.Y. Jeong, Abnormal drop in electrical resistivity with impurity doping of singlecrystal Ag, Sci. Rep. 4 (2014) 5450, https://doi.org/10.1038/srep05450. [28] G. Ding, C. Clavero, D. Schweigert, M. Le, Thickness and microstructure effects in the optical and electrical properties of silver thin films, AIP Adv. 5 (11) (2015) 117234, https://doi.org/10.1063/1.4936637.
4. Conclusions To grow single-crystalline Ag film on Al2O3 substrate that had better adhesion to the substrate, we used a ZnO buffer layer based on crystallographic structure and EADM predictions. The ZnO film showed optimized growth under the 500 °C condition on the Al2O3 substrate, and was aligned epitaxially along the c-axis with grain sizes of ca. 40–50 nm. The growth of Ag film on 5 nm ZnO was optimized at 400 °C, and exhibited an ultraflat surface with an RMS value of 0.73 nm. The XRD peak intensity of the as-grown 100-nm-thick Ag film was five-fold larger than that of the Al2O3 single-crystal substrate, and its EBSD image indicated crystallographic alignment in the [1 1 1] direction. The resistivity of the single-crystalline Ag film was only about half that of the polycrystalline Ag film, and the reflectivity of the single-crystalline film was enhanced. Besides the choice of a ZnO buffer layer, the remarkable improvement in the quality of the Ag film is attributable to the modified sputtering system with a single-crystal Ag target, which enables 2- and 4-inch wafer-scale growth via the stacking of atomic layers. Thus, the results of this study provide a route to simple fabrication of materials requiring ultrahigh adhesion, anti-oxidative stability, and good optical properties for nano-patterning of Ag. CRediT authorship contribution statement Bo-Gwang Jung: Investigation, Writing - original draft, Visualization. Miyeon Cheon: Writing - review & editing, Visualization, Data curation. Su Jae Kim: Formal analysis, Data curation. Alexander Gliserin: Investigation, Methodology. Soo Hoon Chew: Investigation, Methodology. Chae Ryong Cho: Conceptualization, Validation. Seong-Gon Kim: Conceptualization, Formal analysis. Young Hee Lee: Conceptualization, Writing - review & editing. Seungchul Kim: Conceptualization, Data curation, Writing review & editing. Se-Young Jeong: Conceptualization, Data curation, Writing - review & editing, Supervision. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (No. NRF2017R1A2B3011822), (No. NRF-2017R1C1B2006137) and supported by the Institute for Basic Science (IBS-R011-D1). Appendix A. Supplementary material Supplementary data to this article can be found online at https:// doi.org/10.1016/j.apsusc.2020.145705. References [1] E. Ozbay, Plasmonics merging photonics and electronics at nanoscale dimensions, Science 311 (5758) (2006) 189–193, https://doi.org/10.1126/science.1114849. [2] J.A. Schuller, E.S. Barnard, W. Cai, Y.C. Jun, J.S. White, M.L. Brongersma, Plasmonics for extreme light concentration and manipulation, Nat. Mater. 9 (3)
7