Letter pubs.acs.org/NanoLett
Encapsulated Silver Nanoparticles Can Be Directly Converted to Silver Nanoshell in the Gas Phase Peipei Yang, Yong Xu, Lei Chen, Xuchun Wang, Baohua Mao, Zhongzhi Xie, Sui-Dong Wang, Feng Bao, and Qiao Zhang* Jiangsu Key Laboratory for Carbon-Based Functional Materials & Devices, Institute of Functional Nano and Soft Materials (FUNSOM), and Collaborative Innovation Centre of Suzhou Nano Science and Technology, SWC for Synchrotron Radiation Research, Soochow University, Suzhou 215123, P. R. China S Supporting Information *
ABSTRACT: We report, for the first time, that an encapsulated silver nanoparticle can be directly converted to a silver nanoshell through a nanoscale localized oxidation and reduction process in the gas phase. Silver can be etched when exposed to a mixture of NH3/O2 gases through a mechanism analogous to the formation of aqueous Tollens’ reagent, in which a soluble silver−ammonia complex was formed. Starting with Ag@resorcinol-formaldehyde (RF) resin core−shell nanoparticles, we demonstrate that RF-core@Ag-shell nanoparticles can be prepared successfully when the etching rate and RF thickness were well controlled. Due to the strong surface plasmon resonance (SPR) coupling effect among neighboring silver nanoparticles, the RF@Ag nanoparticle showed great SPR and SERS performance. This process provides a general route to the conversion of Ag-core to Ag-shell nanostructures and might be extended to other systems. KEYWORDS: Silver nanoshell, core−shell nanostructure, Tollens’ reagent, surface plasmon resonance, SERS
T
thus highly desired to study both the synthesis and the properties of silver nanoparticles in the gas phase systems, which may shed some new light on the practical applications of silver nanomaterials. However, it is very difficult to synthesize metallic nanoshell through a gas phase reaction because the dielectric core has to be supported on a substrate. Only a partially coated nanostructure could be obtained when metal atoms were deposited from the outer surface. For example, Xia and co-workers have deposited gold half-shell on silica nanoparticle using a sputter.24 To synthesize a fully covered metal nanoshell on the dielectric core, a plausible pathway might be that metal ions were encapsulated in a dielectric shell first, followed by the isotropic diffusion and reduction of metal ions on the surface. However, the encapsulation of metal ions would be even more difficult. Due to the large surface area and high surface energy, the stability, such as thermal-, photo-, and chemical-stability, of nanomaterials is generally lower than that of their bulk counterparts.25−28 Nanoparticles tend to coagulate or “reshape” their morphology in harsh condition or in practical applications, resulting in the loss of their superior properties.29 The chemical stability of nanoparticles is another important issue. For instance, the chemical stability of bulk silver is relatively high, while colloidal silver nanoparticles can be easily damaged by mild oxidants, such as hydrogen peroxide,30,31 halide ions with oxygen,32 Fe3+,33 NH3/O2,34,35 and so forth.
he development of nanomaterials with better performance, higher stability, and lower cost that can be used in practical applications has been the Holy Grail for materials scientists. Among various nanomaterials, silver nanostructures have attracted much attention because of their unique surface plasmon resonance (SPR) property and the potential applications in diverse fields, including surface-enhanced Raman scattering (SERS),1−4 biological sensing,5−7 catalysis,8−10 photonic crystal,11 and so on. Since the SPR property of silver nanoparticles strongly depends on their shape and size, much effort has been devoted to developing new synthetic methodologies to control the morphology of silver nanoparticles.12 After several decades of development, although the advanced nanotechnology allows us to design and synthesize silver nanoparticle with shapes varying from spheres to plates,13−15 wires,16,17 cubes,18 and polyhedrons,19 significant challenges still remain, such as low-volume production, complicated synthetic approach, and high cost. Recently, silver nanoshell, a class of nanoparticles that are composed of a dielectric core and a concentric silver layer, has drawn much attention due to its tunable SPR property and the application in many fields, such as SERS and biosensing.20−23 Although much attention has been paid to the colloidal synthesis of silver nanoshells, it is still a great challenge to prepare silver nanoshells through an efficient and cost-effective way.20−23 To date, most of the studies have been focused on colloidal silver nanoparticles. Much less attention has been paid to the gas phase systems, despite the fact that silver nanoparticles have been widely used in devices that work under gas phase. It is © XXXX American Chemical Society
Received: October 25, 2015 Revised: November 20, 2015
A
DOI: 10.1021/acs.nanolett.5b04328 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
dots were observed on the surface of RF shell (Figure 1c). Careful examination showed that the silver core in the center was partially removed. Two weeks later, silver core was removed completely while a black shell was formed on the RF surface. At the same time, a hollow cavity can be clearly observed in the TEM image (Figure 1d). Energy dispersive Xray (EDX) elemental mapping of a single particle (Figure 1e− h) clearly confirms that the outer shell is composed of silver species. It is worth pointing out that the as-prepared core−shell nanostructure is very stable. The size of silver particles on the surface of RF spheres kept almost the same when the samples were stored in air for two months. When the shell thickness was increased to over 100 nm, an RF@Ag core−satellite structure was obtained because there was not enough silver that can form a dense coating on the surface (Figure S1). To reveal the composition of silver species, a series of experiments were conducted. First, the product was characterized by using the high-resolution TEM (HRTEM). As shown in Figure 2a, the interplanar spacing of the lattice fringe is 0.24 nm, which is in great agreement with the (111) facet of metallic silver with face-centered-cubic (fcc) structure. It is surprising that Ag2O has not been observed as silver nanoparticle can be oxidized easily in air. The product was then examined by the X-ray photoelectron spectroscopy (XPS) study (Figure S3 and Figure 2b). The spectrometer was calibrated with one reference point (C 1s at 284.4 eV) to minimize the error due to the drift of the spectrometer caused by the charging effects. Two peaks, 374.1 and 368.1 eV, can be observed in the Ag 3d spectrum (Figure 2b), which can be assigned to the Ag 3d3/2 and 3d5/2 of metallic silver, respectively.37,38 The metallic nature of the product was further confirmed by its SPR property and SERS performance. It is well-known that the so-called “plasmonic coupling” effect will appear when two metallic nanoparticles get close enough to each other.39 In this transformation process, there were two distinguishable features: the increase in both nanoparticle size and number, and the decrease in separation between two neighboring silver nanoparticles. Because the silver shell is quite dense, a red-shift in the SPR peak will be expected if the shell was composed of metallic silver. As shown in Figure 2c, the original Ag@RF core−shell nanostructure has a sharp extinction peak at ∼450 nm. With the reaction time prolonged, the separation between neighboring silver nanoparticles decreased gradually, resulting in the increase of the dipole coupling, which is responsible for the red-shifting and broadening of the plasmon resonance bands.40 The SPR peak could shift to over 700 nm. Additionally, because only metallic silver can be used for the surface enhanced Raman scattering (SERS) application due to its SPR property, a SERS measurement was carried out to confirm the metallic nature of silver shell. R6G molecules were used as the target molecules. As shown in Figure 2d, no signal was detected when naked silicon wafer was used as the substrate ([R6G] = 10−4 M). When RF@Ag core−shell nanostructures were used as the substrate, characteristic peaks were observed. The limit of detection is about 10−7 M. Although we cannot rule out the possibility that some surface atoms were oxidized in air, the above results verified that the majority of the product should be metallic silver. The great chemical stability of the as-obtained silver nanoparticles against the oxidation might be attributed to the reductive surface of RF spheres, which can effectively reduce the oxidized species to metallic silver.
Inspired by the studies in the colloidal systems, we demonstrate here that silver nanoparticles can be completely etched away in a mixture of ammonia and oxygen gas with a mechanism analogous to the formation of Tollens’ reagent in the liquid phase, in which a soluble silver−ammonia complex, [Ag(NH3)2]+, was formed. The complex can migrate in the gas phase, resulting in the “decomposition” of silver nanoparticles. When the etching process and the following reduction process were confined in a limited nanoscale space, silver nanoparticle could be converted to a silver shell structure. By carefully tuning the reaction rate and the shell thickness, core−satellite or core−shell nanostructures can be obtained. This process is a general process that can happen in other core−shell nanostructures with a silver core. Figure 1a illustrated the transformation process in which Ag@RF nanoparticles were converted to RF@Ag core−shell
Figure 1. (a) Schematic illustration of the chemical transformation process of Ag@RF core−shell nanostructures: Ag@RF nanoparticles were etched by oxygen in the presence of gaseous ammonia. Due to the reductive surface of RF shell, the oxidized silver was then reduced to form a silver shell on the surface of RF resin. (b−d) TEM images showing the evolution process: (b) the original Ag@RF nanoparticles; (c) partially etched and reduced nanostructure with some metallic silver dots on the surface; (d) the final product in which silver core was completely etched away and a silver shell was formed on the surface. (e−h) The HAADF image (e) and the elemental mapping of the Ag@ RF core−shell nanostructure: (f) C; (g) O; and (h) Ag. The scale bars in (e−h) are 20 nm.
nanostructures with a hollow cavity in the center. In the presence of ammonia gas, silver core could be etched gradually by oxygen. Because the polymer shell is reductive, the oxidized silver ions can be reduced to form silver dots on the surface of RF resin. With the reaction time prolonged, silver core could be completely etched away, and a thin layer of silver shell was formed. The transformation process was monitored by the TEM characterization. The Ag@RF resin core−shell nanostructures were prepared according to a previous report in which resorcinol, formaldehyde, and silver nitrate were mixed and refluxed for 30 min in the presence of ammonia hydroxide.36 As shown in Figure 1b, the core size of the product is around 45 nm and the shell thickness is about 30 nm. After dried and exposed in air for 3 days, some small black B
DOI: 10.1021/acs.nanolett.5b04328 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 2. (a) HRTEM image showing that interplanar spacing of the lattice fringes is 0.24 nm, which is in good agreement with the (111) spacing of metallic silver; (b) the Ag 3d spectrum of the RF@Ag core−shell nanostructure; (c) the UV−vis spectra showing the SPR peak of the Ag@RF composite red-shifted during the transformation process. From left to right: UV−vis spectra of samples after stored in air for (black) 0 day; (red) 3 days; (blue) 1 week; and (pink) 2 weeks. (d) Raman spectra of R6G adsorbed on bare silicon substrate ([R6G] = 10−4 M, black line) and on RF@ Ag core−shell nanostructure ([R6G] = 10−4 M, red line; 10−5 M, purple line; 10−6 M, green line; and 10−7 M, yellow line).
The mechanism of this transformation process has been carefully studied. The role of ammonia gas was examined first. To eliminate the residual of ammonia, the obtained Ag@RF nanoparticles were thoroughly washed with water for more than 15 times and dried in a vacuum condition. No etching process could be observed even when the product was stored in air for more than three months, indicating the critical role of ammonia gases. In the presence of abundant ammonia gas, the etching process happened very fast. As shown in Figure S4, silver nanoparticles could be etched away within 2 h when Ag@ RF nanoparticles were put on the top of a cup of concentrated ammonia solution (28%). Because the reaction is so fast, no silver shell could be obtained on the surface of RF shell. Only some silver nanoparticles were left on the hollow RF spheres. Detailed examination showed that silver nanoshell could be obtained when the sample was stored in air at room temperature with the concentration of ammonia in the range of 10−100 ppm. Oxygen is also critical in this process. When the Ag@RF nanoparticles were stored in an inert atmosphere (ultrapure nitrogen gas), no etching process could be observed even when ammonia gas was injected into the system (the final volume percentage of ammonia was 1%). Combining the above results, a plausible mechanism has been proposed: since ammonia hydroxide was used to catalyze
the polymerization process of resorcinol−formaldehyde, there were some residuals of ammonia in the Ag@RF nanoparticles. When exposed to air, silver nanoparticles could be oxidized to form Ag2O on the surface, which was then reacted with ammonia to form [Ag(NH3)2]+ complex, according to eq 1. Ag + O2 + 8NH3·H 2O → 4[Ag(NH3)2 ]+ + 4OH− + 6H 2O (1)
The complexes could diffuse out and were reduced by the reductive surface of RF shell (either the R−OH group, or the residual formaldehyde or resorcinol). The migration of the complex ions has been confirmed by mixing Ag@RF nanoparticles with pure RF nanoparticles. When the mixture was dried and exposed in air, small silver nanoparticles could be observed on the surface of both Ag@RF and pure RF nanoparticles after 2 days (Figure S5). Two weeks later, the silver core has been completely removed. A dense silver coating formed on the original RF surface, while fewer and smaller silver nanoparticles could be observed on the surface of pure RF nanoparticles, suggesting that most of the silver−ammonia complexes have been reduced on the original location. On the basis of the mechanism proposed above, the most important intermediate is the [Ag(NH3)2]+ complex. Therefore, we used the Tollens’ reagent to test the hypothesis. The C
DOI: 10.1021/acs.nanolett.5b04328 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters Tollens’ reagent was prepared by adding ammonia hydroxide solution dropwise into a silver nitrate solution until a clear solution was obtained. The Ag@RF nanoparticle with a core size ∼40 nm and a shell thickness of ∼100 nm was used as the template (Figure 3a). A freshly prepared Tollens’ reagent was
Figure 4. TEM images showing the evolution process of (a−c) AgNW@SiO2 core−shell and (d−f) AgNW@SiO2@RF core−shell− shell nanostructure in the presence of both oxygen and ammonia gas. (a) The original AgNW@SiO2 nanostructure and after exposed in air for (b) 3 days and (c) 2 weeks; (d) the original AgNW@SiO2@RF nanostructure and after exposed in air for (e) 3 days and (f) 2 weeks.
removed, and silver nanoparticles were deposited back to the surface of silica shell. The reduction of silver−ammonia complex on the silica surface can be explained by the nucleophilic substitution and electrophilic addition mechanism.41 In another case, the AgNW@SiO2 core−shell nanostructure was coated with a layer of RF resin by using an extended Stöber method in which the polymerization between resorcinol and formaldehyde was catalyzed by ammonia hydroxide (Figure 4d).42 An interesting phenomenon is that silver nanoparticles tended to deposit on the surface of RF rather than the interface of SiO2/RF (Figure 4e,f). The elemental mapping results are shown in Figure S7, from which a layer of silver can be clearly observed on the outer surface of RF shell. In conclusion, we have described the first example of converting silver nanoparticles to silver nanoshell through a nanoscale localized oxidative etching and reduction process in the gas phase. It is believed that ammonia gas played an important role in this transformation process by helping form the silver−ammonia complexes, which is similar to what happened in the liquid phase. The as-prepared composite structure showed great SPR and SERS performance. This transformation process is a general process that can happen in other encapsulated silver nanostructures. An interesting aspect of our system is that this approach might be extended to other metal systems when appropriate oxidants and other parameters were presented. This research not only can be helpful in realizing the stability issue of nanomaterials in the gas phase but also be useful in developing novel methodologies for making nanomaterials with desired structure and properties.
Figure 3. (a,c) TEM images and (d) SEM image showing the preparation of a silver shell on the RF surface by treating Ag@RF nanoparticles with the Tollens’ reagent. (a) The original Ag@RF nanoparticle; (b) after treated with the Tollens’ reagent for 5 min; and (c,d) after dried and exposed in air for 48 h.
injected into the Ag@RF aqueous solution under sonication. About five seconds later, the color of the reaction system changed from yellow to dark brown, indicating the reduction of silver. TEM image (Figure 3b) shows that the RF surface has been decorated with some small silver nanoparticles (∼8 nm) after being sonicated for 5 min. It is worth noting that no obvious change could be observed if the Ag@RF nanoparticles were stored in the Tollens’ reagent (up to 5 days), suggesting that no dense silver shell could be obtained by simply treating the RF surface with the Tollens’ reagent. After separated from the solution, the composite was dried and exposed in air. A dense silver shell could be observed after 48 h (Figure 4c,d). Elemental mapping images also confirmed that silver is uniformly distributed on the surface (Figure S6). The asprepared Ag@RF@Ag nanostructure shows similar UV−vis property as that of RF@Ag nanoparticles with a broad SPR peak around 700 nm, suggesting the formation of metallic silver shell. It is worth pointing out that if the composite was washed thoroughly after sonication, no change could be observed after two months, further confirming the proposed mechanism. This transformation process is a general process that can happen in other core−shell nanostructures with a silver core. Here, we used AgNW@SiO2 and AgNW@SiO2@RF nanostructures as the examples. Silver nanowires (NWs) were first coated with a silica layer through a classical Stöber method, in which the hydrolysis of tetraethoxysilane (TEOS) was catalyzed by ammonia hydroxide in an ethanol/water mixture to form a coaxial silica layer on the surface of silver nanowire.35 After dried and exposed in air, similar phenomena were observed. As shown in Figure 4a−c, silver nanowires were gradually
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.5b04328. Materials and methods, additional TEM images, elemental mapping images, and XPS data (PDF) D
DOI: 10.1021/acs.nanolett.5b04328 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
■
(24) Lu, Y.; Xiong, H.; Jiang, X. C.; Xia, Y. N.; Prentiss, M.; Whitesides, G. M. J. Am. Chem. Soc. 2003, 125, 12724−12725. (25) Pastoriza-Santos, I.; Liz-Marzan, L. M. J. Mater. Chem. 2008, 18, 1724−1737. (26) Burda, C.; Chen, X. B.; Narayanan, R.; El-Sayed, M. A. Chem. Rev. 2005, 105, 1025−1102. (27) Goldstein, A. N.; Echer, C. M.; Alivisatos, A. P. Science 1992, 256, 1425−1427. (28) Giersig, M.; Ung, T.; Liz-Marzan, L. M.; Mulvaney, P. Adv. Mater. 1997, 9, 570−575. (29) Zhang, Q.; Lee, I.; Ge, J. P.; Zaera, F.; Yin, Y. D. Adv. Funct. Mater. 2010, 20, 2201−2214. (30) Zhang, Q.; Li, N.; Goebl, J.; Lu, Z. D.; Yin, Y. D. J. Am. Chem. Soc. 2011, 133, 18931−18939. (31) Gao, C. B.; Lu, Z. D.; Liu, Y.; Zhang, Q.; Chi, M. F.; Cheng, Q.; Yin, Y. D. Angew. Chem., Int. Ed. 2012, 51, 5629−5633. (32) Rycenga, M.; Cobley, C. M.; Zeng, J.; Li, W. Y.; Moran, C. H.; Zhang, Q.; Qin, D.; Xia, Y. N. Chem. Rev. 2011, 111, 3669−3712. (33) Wiley, B.; Sun, Y. G.; Xia, Y. N. Langmuir 2005, 21, 8077−8080. (34) Hunyadi, S. E.; Murphy, C. J. J. Phys. Chem. B 2006, 110, 7226− 7231. (35) Yin, Y. D.; Lu, Y.; Sun, Y. G.; Xia, Y. N. Nano Lett. 2002, 2, 427−430. (36) Yang, P.; Xu, Y.; Chen, L.; Wang, X.; Zhang, Q. Langmuir 2015, 31, 11701−11708. (37) Lai, Y. K.; Zhuang, H. F.; Xie, K. P.; Gong, D. G.; Tang, Y. X.; Sun, L.; Lin, C. J.; Chen, Z. New J. Chem. 2010, 34, 1335−1340. (38) Sumesh, E.; Bootharaju, M. S.; Pradeep, A. T. J. Hazard. Mater. 2011, 189, 450−457. (39) Jain, P. K.; El-Sayed, M. A. Chem. Phys. Lett. 2010, 487, 153− 164. (40) Zhang, Q.; Ge, J. P.; Goebl, J.; Hu, Y. X.; Sun, Y. G.; Yin, Y. D. Adv. Mater. 2010, 22, 1905−1909. (41) Kim, Y. H.; Lee, D. K.; Cha, H. G.; Kim, C. W.; Kang, Y. S. J. Phys. Chem. C 2007, 111, 3629−3635. (42) Liu, J.; Qiao, S. Z.; Liu, H.; Chen, J.; Orpe, A.; Zhao, D. Y.; Lu, G. Q. Angew. Chem., Int. Ed. 2011, 50, 5947−5951.
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Author Contributions
P.Y. and Y.X. contributed equally to this work. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (21401135) and the Natural Science Foundation of Jiangsu Province (BK20140304). This project is funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD). Y.X. acknowledges the Collaborative Innovation Center of Suzhou Nano Science and Technology (Nano-CIC) for the fellowship of “Collaborative Academic Training Program for Post-doctoral Fellows” and the postdoctoral starting funding support from Soochow University.
■
REFERENCES
(1) Jackson, J. B.; Halas, N. J. Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 17930−17935. (2) Abalde-Cela, S.; Ho, S.; Rodriguez-Gonzalez, B.; Correa-Duarte, M. A.; Alvarez-Puebla, R. A.; Liz-Marzan, L. M.; Kotov, N. A. Angew. Chem., Int. Ed. 2009, 48, 5326−5329. (3) Lee, J. H.; You, M. H.; Kim, G. H.; Nam, J. M. Nano Lett. 2014, 14, 6217−6225. (4) Peng, S.; Lei, C. H.; Ren, Y.; Cook, R. E.; Sun, Y. G. Angew. Chem., Int. Ed. 2011, 50, 3158−3163. (5) Anker, J. N.; Hall, W. P.; Lyandres, O.; Shah, N. C.; Zhao, J.; Van Duyne, R. P. Nat. Mater. 2008, 7, 442−453. (6) Jain, P. K.; Huang, X. H.; El-Sayed, I. H.; El-Sayed, M. A. Acc. Chem. Res. 2008, 41, 1578−1586. (7) Zhang, Y.; Zhu, C. F.; Zhang, L.; Tan, C. L.; Yang, J.; Chen, B.; Wang, L. H.; Zhang, H. Small 2015, 11, 1385−1389. (8) Christopher, P.; Xin, H. L.; Linic, S. Nat. Chem. 2011, 3, 467− 472. (9) Liu, X. H.; Ma, J. G.; Niu, Z.; Yang, G. M.; Cheng, P. Angew. Chem., Int. Ed. 2015, 54, 988−991. (10) Deng, Z.; Chen, M.; Wu, L. J. Phys. Chem. C 2007, 111, 11692− 11698. (11) Wang, W.; Asher, S. A. J. Am. Chem. Soc. 2001, 123, 12528− 12535. (12) Xia, Y. N.; Xiong, Y. J.; Lim, B.; Skrabalak, S. E. Angew. Chem., Int. Ed. 2009, 48, 60−103. (13) Jin, R. C.; Cao, Y. C.; Hao, E. C.; Metraux, G. S.; Schatz, G. C.; Mirkin, C. A. Nature 2003, 425, 487−490. (14) Jin, R. C.; Cao, Y. W.; Mirkin, C. A.; Kelly, K. L.; Schatz, G. C.; Zheng, J. G. Science 2001, 294, 1901−1903. (15) Huang, X.; Zeng, Z. Y.; Bao, S. Y.; Wang, M. F.; Qi, X. Y.; Fan, Z. X.; Zhang, H. Nat. Commun. 2013, 4, 1444−1451. (16) Sun, Y. G.; Gates, B.; Mayers, B.; Xia, Y. N. Nano Lett. 2002, 2, 165−168. (17) Zhu, L. F.; Shen, X. S.; Zeng, Z. Y.; Wang, H.; Zhang, H.; Chen, H. Y. ACS Nano 2012, 6, 6033−6039. (18) Sun, Y. G.; Xia, Y. N. Science 2002, 298, 2176−2179. (19) Tao, A.; Sinsermsuksakul, P.; Yang, P. D. Angew. Chem., Int. Ed. 2006, 45, 4597−4601. (20) Jackson, J. B.; Halas, N. J. J. Phys. Chem. B 2001, 105, 2743− 2746. (21) Halas, N. J. MRS Bull. 2005, 30, 362−367. (22) Jankiewicz, B. J.; Jamiola, D.; Choma, J.; Jaroniec, M. Adv. Colloid Interface Sci. 2012, 170, 28−47. (23) Jiang, Z. J.; Liu, C. Y. J. Phys. Chem. B 2003, 107, 12411−12415. E
DOI: 10.1021/acs.nanolett.5b04328 Nano Lett. XXXX, XXX, XXX−XXX
