Communication pubs.acs.org/JACS
Transformative Heterointerface Evolution and Plasmonic Tuning of Anisotropic Trimetallic Nanoparticles Mouhong Lin, Gyeong-Hwan Kim, Jae-Ho Kim, Jeong-Wook Oh, and Jwa-Min Nam* Department of Chemistry, Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul 08826, South Korea S Supporting Information *
anisotropic configuration not only allows the integration and coupling of the characteristic SPRs of individual metal components but also enables unique spatial organization of the sub-NPs and their plasmonic propagation.8,16 The design and synthesis of multicomponent NPs are often complicated and challenging due to various growth modes, many possible multiple-component arrangements and particle stability issue. It is postulated the surface energies and interfacial energies involved at the heterointerface are the driving forces in multicomponent hybrid formation.17 If the transformation and evolution of these heterointerface are better understood, the growth patterns and structure control of MCPs may be predictable and controllable. Here, we developed a transformative heterointerface evolution (THE) method for the controlled synthesis of gold@copper− silver (G@CS) and copper−gold−silver (CGS) MAPs with the aid of polyethylenimine (PEI), polyvinylpyrrolidone (PVP) and galvanic replacement reaction, and show the heterojunctions and plasmonic properties of two series of well-defined MAPs can be tuned in a controlled manner (Figures 1, 2 and S1). We demonstrated the critical role of seed crystal structure in multimetallic hybrid structure growth, and importantly, the size, composition order and morphology of a heterojunction can be tuned in these MAPs, giving rise to the exceptional designability of structure and property. In our scheme, bimetallic template engineering is critical to the subsequent formation of well-defined MAPs through galvanic replacement reaction. As illustrated in Figure 1a, ∼10 nm gold single-crystalline seeds (GSS) and gold icosahedral twin seeds (GTS) were employed to form gold−copper core−shell (GCCS) and gold−copper tip-taper (GCTT) NPs, respectively. In a typical experiment, GSSs or GTSs were formed and injected into the growth solution containing copper(II) chloride, PEI and ascorbic acid (AA). The amine-rich carbon chain backbone of PEI is a good capping ligand for stabilizing copper NPs.18 It readily protects exposed copper surface and remains highly water dispersible. From Cs-corrected high resolution transmission electron microscopy (HRTEM), single crystallinity was observed for GSSs (Figures 1b, S2) whereas the presence of twin was distinguished in GTSs (Figures 1e, S2). Fast Fourier transform (FFT) patterns that convert HRTEM images into the reciprocal lattice patterns further reveal the copper shell epitaxially grew from the GSS and largely maintained a single-crystalline structure (Figure 1c,d). Slip planes (Figure 1d), stemming from dislocations between the planes of copper atoms, were observed due to the strain within the copper shell. Regarding GCTT, Figure
ABSTRACT: Multicomponent nanoparticles that incorporate multiple nanocrystal domains into a single particle represent an important class of material with highly tailorable structures and properties. The controlled synthesis of multicomponent NPs with 3 or more components in the desired structure, particularly anisotropic structure, and property is, however, challenging. Here, we developed a polymer and galvanic replacement reaction-based transformative heterointerface evolution (THE) method to form and tune gold−copper−silver multimetallic anisotropic nanoparticles (MAPs) with welldefined configurations, including structural order, particle and junction geometry, giving rise to extraordinarily high tunability in the structural design, synthesis and optical property of trimetallic plasmonic nanoantenna structures. MAPs can easily, flexibly integrate multiple surface plasmon resonance (SPR) peaks and incorporate various plasmonic field localization and enhancement within one structure. Importantly, a heteronanojunction in these MAPs can be finely controlled and hence tune the SPR properties of these structures, widely covering UV, visible and near-infrared range. The development of the THE method and new findings in synthesis and property tuning of multicomponent nanostructures pave ways to the fabrication of highly tailored multicomponent nanohybrids and realization of their applications in optics, energy, catalysis and biotechnology.
C
olloidal multicomponent nanoparticles (MCPs)1−3 are beneficial in that these particles can utilize the advantageous physicochemical properties of all the components within one nanoparticle (NP), and more finely engineered structures can be designed in this manner.4,5 Because of surface plasmon propagation, such synergistic effects resulting from the electromagnetic communication at the solid-state heterojunction are particularly significant in anisotropic MCPs composed of multiple plasmonic subnanoparticles (sub-NPs). The high plasmonic tunability in multimetallic anisotropic nanoparticles (MAPs) render them as promising candidates in metallic structures6−8 that facilitate diverse plasmonic effects in solar cell,9 catalysis,10 sensing,11 photochemistry12 and nanoantenna.13 In particular, the plasmonic effects of metal NPs have been explored to solve the low light manipulation efficiency in light harvesting, transportation and conversion processes. Such effects are, however, limited to specific wavelengths of light due to the narrow surface plasmon resonance (SPR) of metal NPs.14,15 A multicomponent © 2017 American Chemical Society
Received: May 5, 2017 Published: July 19, 2017 10180
DOI: 10.1021/jacs.7b04202 J. Am. Chem. Soc. 2017, 139, 10180−10183
Communication
Journal of the American Chemical Society
1g evidences the heteroepitaxial growth of copper domain from one side of the GTS surface, and three distinct domains were observed in copper. FFT patterns indicate these three domains are single-crystalline and the two side domains (FFT 1 and 3 in Figure 1g) are showing symmetric lattice patterns, characteristic of cyclic twin. The angle between these two grains is ∼120°, and there is no shared ⟨111⟩ plane, which indicates the two side domains correspond to two nonadjacent grains in a 5-fold twin (for more HRTEM analysis and explanations, see Figure S3). These results show copper selectively grew from one of the 5-fold twin edges of a GTS and epitaxially formed a 5-fold twined body. For the strategic synthesis of multimetallic structures with transformative heterointerface, the galvanic replacement reactions were performed by mixing an aqueous solution of AgNO3 and PVP with GCCS or GCTT templates at room temperature for 15 min, respectively. The replacement of copper by silver is a two-electron reaction, in which two portions of silver will be deposited when one portion of copper is consumed. Therefore, the sizes of copper and silver components can be tuned simultaneously through altering the concentration of AgNO3 ([Ag+]) to control the extent of galvanic replacement reaction. Figure 2a,c shows two series of MAPs synthesized with different [Ag+] (Figure S4, S5). Structural order of the three components was well-defined and controlled in G@CS and CGS MAP series. The spatial distributions of gold, copper and silver in MAPs were characterized by energy-dispersive X-ray spectroscopy (EDX) mapping. As [Ag+] increased, the size of copper domain gradually decreased before finally disappearing, and instead a newly emerged silver domain was observed. The reaction was also monitored by UV−vis extinction spectrum and solution color (Figure 2b,d). The results suggest the optical properties of these nanostructures are tunable and solution color change was visible. As reaction preceded, the copper SPR peaks (∼570 nm) decreased with increasing [Ag+], and only the silver SPR peak (∼410 nm) was remained after the completion of galvanic replacement reaction. We further studied the roles of PVP and PEI in controlling MAP morphology and heterointerface. The zeta-potentials of GCCS and GCTT templates dispersed in water were 27.0 ± 0.7 mV and 22.4 ± 0.9 mV, respectively, and the values dropped to 14.6 ± 0.4 mV and 14.9 ± 0.5 mV, respectively, after the addition of PVP. Because PVP has a nearly neutral charge, the large ζpotential decrease upon PVP addition suggests PVP may partially exchange with PEI during the galvanic replacement reaction (Figure 3a). When [PVP] is low, the high surface energy of silver sub-NP lead to a spherical shape, whereas the copper surface is well protected with PEI along with minimal replacement of PEI with PVP, resulting in a narrow junction between copper surface and silver sub-NPs (Figures 3b-I,c-I and S8). As [PVP] increases, the surface energy of the silver sub-NP is reduced, and the exchange between PEI and PVP promotes the expansion of the heterojunction (Figure 3a). Regarding the G@CS MAP series (Figures 3b and S6), the heterojunction expansion direction is tangent to the particle longitudinal axis. The junction width expands with increase in [PVP], whereas the configuration of subNPs remains similar. As a result, its structure gradually shifts from a small-junction dimer (G@CS-1-1) to a thick-junction dimer structure (G@CS-1-4). In contrast, regarding the CGS MAP series (Figures 3c and S7), the heterojunction expansion direction moves along the longitudinal axis of the particle due to the presence of a gold sub-NP at the junction. Interestingly, a dumbbell-like trimer structure (CGS-1-1) was formed at low [PVP], in which the GTS serves as a “bridge” between copper and
Figure 1. (a) Schematic illustration of the crystal structure engineering in bimetallic templates. HRTEM images of (b) a gold single-crystalline seed (GSS), (c, d) a gold−copper core−shell (GCCS) NP, (e) two representative gold icosahedral twin seeds (GTSs), and (f, g) a gold− copper tip-taper (GCTT) NP, respectively. For both GCCS and GCTT, identical growth conditions were used except the difference of gold seeds. Scale bars are 5 nm.
Figure 2. Schematic illustration of galvanic replacement reaction using (a) GCCS and (c) GCTT NP as templates. HRTEM images and corresponding EDX mapping images of two different bimetallic templates, as well as their galvanic replacement products synthesized by reacting AgNO3 with (a) GCCS and (c) GCTT NPs, respectively. Extinction spectra of a series of (b) G@CS and (d) CGS MAPs. Scale bars are 20 nm.
10181
DOI: 10.1021/jacs.7b04202 J. Am. Chem. Soc. 2017, 139, 10180−10183
Communication
Journal of the American Chemical Society
forms a gold−silver heterointerface. Multiple nucleation was not observed because the fresh silver nuclei become advantageous electrode surface comparing with other PEI-capped template surface, resulting in the high-order, well-crystallized structure. Next, we explored the plasmonic properties of these transformative trimetallic nanostructures. Figure 4a,b shows the extinction spectra of G@CS-1 and CGS-1 MAP structures. As [PVP] increased, the corresponding near-infrared (NIR) peaks were gradually blue-shifted, indicating the junction-dependent SPR. Nevertheless, in both G@CS (Figure 4a, inset) and CGS (Figure 4b, inset) cases, the NIR peak wavelength (black dots) exhibited a high linearity with [PVP], indicating the junction morphology and corresponding plasmonic property are highly correlated. The junction controllability in MAPs is of particular importance due to the strong antenna effect in these structures and their high sensitivity to heterojunction geometry.8,19 We found the far-field and near-field properties of these antennas can be finely tuned in MAP structures, and a large part of the sun spectrum, especially in the NIR region, can be covered by these structures (Figures 2b,d, 4a,b). It should be noted the extinction peak intensity remained strong throughout the NIR region for CGS MAPs (Figure 4b). To understand the junction-dependent plasmonic coupling between the sub-NPs of MAPs, threedimensional finite-element method (3D FEM) simulations were conducted. Our initial G@CS model system is composed of a 40 nm GCCS cube and a 25 nm silver NP, whereas the contacted junction between them varies from 5 to 14 nm in diameter (Figure 4c). As the junction width increased, the SPR peak was largely blue-shifted from 1280 to 700 nm, which is consistent with our experimental data (Figure 4a). This result matches also with the previous theoretical and experimental data, and this implies that such low-frequency SPR corresponds to the charge transfer plasmon (CTP) mode.16 Because of the CTP, intense electric fields are localized at the junction area (Figure 4e), which is up to 219-fold higher than the incident light for a 5 nm-junction G@CS MAP (M1). This localization effect is inversely proportional to the junction width. As near-field localization effect increases with junction narrows down (from M4 to M1), the corresponding peak width rapidly broadens both in experimental (Figure 4a) and theoretical (Figure 4c) results. Regarding models M5 to M8, the silver sub-NP was positioned from 4 nm apart to 2 nm apart, in
Figure 3. Heterointerface evolutions in MAPs. (a) Heterointerface engineering strategies with GCCS and GCTT. TEM images of heterojunction-engineered gold@copper−silver (G@CS) (b) and copper−gold−silver (CGS). (c) MAP series show four different junction morphologies (I to IV). Scale bars are 20 and 5 nm in low-magnification and high-magnification TEM images, respectively.
silver sub-NPs. As [PVP] increased, the “bridge” length shortened accordingly, and the gold was eventually buried by silver (CGS-14). Because of surface protection of the copper surface of GCCS NPs by PEI, silver slowly nucleates on the copper surface. After the initial silver nucleation site is formed, this silver NP becomes a preferential site for further silver structure growth because other copper surface is still well protected with PEI. The exposed gold tip in GCTT templates works as a “nano-electrode” (schematically illustrated in Figure 2c) for silver’s under potential deposition. Instead of the direct electron transfer from copper to silver, the galvanic current passes through the gold NP and
Figure 4. Plasmonic heterojunction-engineered optical properties. Extinction spectra of (a) G@CS-1 and (b) CGS-1 MAPs synthesized by varying PVP concentrations. Insets show the effect of PVP concentration on NIR peak wavelength (solid black dots) or peak intensity (hollow red bars). The extinction cross sections of (c) the junction-engineered G@CS (models M1 to M4) and (d) CGS (models M5 to M8) MAPs in the NIR region. (e) Corresponding electric field distributions (left) and magnified junction areas (right) at their respective charge transfer plasmon (CTP) peak wavelengths. (f) Volume integral of square of electric field (E2) from the 2 nm-periphery of each structure (∫ Eperiphery2, gray bars) and the defined junction area (∫ Ejunction2, black bars), shown in Figure 4e. Scale bars are 5 nm. 10182
DOI: 10.1021/jacs.7b04202 J. Am. Chem. Soc. 2017, 139, 10180−10183
Journal of the American Chemical Society
■
contact with the copper sub-NP and overlapped with copper subNP (Figure 4d). Other than M4 (the thick junction case) and M8 (the overlapped case), the electromagnetic field was strongly localized within the junction area (Figure 4e). The optical energy, namely the power stored with the antenna, is proportional to the integral of square of the local electric field (E2). Quantitative analysis was performed by volume integral of E2 on these antenna structures. As shown in Figure 4f, the gray bars (∫ Eperiphery2) display the volume integral of E2 over the 2 nm peripheral area of models M1 to M8 (peripheral field; Figure S9), and the black bars (∫ Ejunction2) represent the E2 contribution from their junction regions (junction field). The ∫ Ejunction2 to ∫ Eperiphery2 ratio describes the electric field localization efficiency of an antenna. Overall, the highly localized field in the small junction area was much stronger than the overall field intensity on particle surface for 80 nm gold nanoparticle and 80 nm gold nanorod. From M4 to M1, the ∫ Ejunction2 to ∫ Eperiphery2 ratio rapidly increases from 24% (M4) to 63% (M1), indicative of the importance of the junction size in inducing stronger localized field around the junction. Among M5 to M8, the M7, in which silver and copper sub-NPs are in contact, displayed the strongest field intensity around the junction area. In contrast, the electric field was remarkably decreased when the junction area was buried with silver (Figure 4e,f, M8). M1 shows the strongest localization efficiency in the near-field, but its ∫ Eperiphery2 value (Figure 4f) and far-field peak intensity (Figure 4c) are the lowest. This phenomenon was also observed in the plasmonic nanosnowmen with a conductive nanojunction8 and nanowire-bridged dimer,16 and was theoretically studied in the plasmonic cavities with a conductive junction20 and the rod antenna with a conductive filament.21 As junction narrows down, the SPR peak intensity rapidly damps and the peak width largely broadens, which are due to the energy dissipation at the junction.20 As a result, the near-field properties of these antenna structures are inversely correlated with their farfield properties, which limit their applications. In summary, we developed new material designing and synthetic approaches in forming and tuning trimetallic nanostructures via finely engineering galvanic replacement, PEI and PVP on crystal structure-engineered bimetallic templates. We showed two series of well-defined MAPs (transformative gold@ copper−silver MAP and copper−gold−silver MAP) can be synthesized in a high yield, and their structures and plasmonic properties are controllable. Our approach allows for efficient integration and coupling of characteristic SPR properties of subnanocomponents within a single MAP, showing high potential for extending plasmonic enhancement in various plasmon-enhanced applications. Moreover, nanometer-scale heterojunction engineering capability with three different components, the tunability over the “confined-neck junction” to “open-neck junction” and the fine balancing of far-field and near-field properties in copper−gold−silver MAPs offer opportunity for plasmonic control and enhancement in various metal and semiconductor-based energy applications or diffusion-based catalytic applications. These capabilities can provide new insights and prompt new theoretical and experimental investigations in plasmonics and photochemistry. The strategies and results herein open avenues for the design, synthesis, optical tunability and plasmonic applications of anisotropic multicomponent metal nanostructures.
Communication
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b04202. Experimental details and characterization data (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*
[email protected] ORCID
Jwa-Min Nam: 0000-0002-7891-8482 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS J.-M.N. was supported by the National Research Foundation (NRF) of Korea (2016R1A2A1A05005430) and BioNano Health-Guard Research Center funded by the Ministry of Science, ICT & Future Planning (MSIP) of Korea as Global Frontier Project (H-GUARD_2013M3A6B2078947). J.-M.N. also acknowledges the support from the Pioneer Research Center Program (2012-0009586) through the NRF of Korea funded by the Ministry of Science, ICT and Future Planning. M. Lin acknowledges the support from China Scholarship Council (CSC, 201308440259).
■
REFERENCES
(1) Buck, M. R.; Bondi, J. F.; Schaak, R. E. Nat. Chem. 2012, 4, 37. (2) Zhou, B.; Shi, B. Y.; Jin, D. Y.; Liu, X. G. Nat. Nanotechnol. 2015, 10, 924. (3) Sharma, S.; Yamazoe, S.; Ono, T.; Kurashige, W.; Niihori, Y.; Nobusada, K.; Tsukuda, T.; Negishi, Y. Dalton Trans. 2016, 45, 18064. (4) Mokari, T.; Sztrum, C. G.; Salant, A.; Rabani, E.; Banin, U. Nat. Mater. 2005, 4, 855. (5) Wang, H.; Chen, L.; Feng, Y.; Chen, H. Acc. Chem. Res. 2013, 46, 1636. (6) Gramotnev, D. K.; Bozhevolnyi, S. I. Nat. Photonics 2010, 4, 83. (7) Schuller, J. A.; Barnard, E. S.; Cai, W. S.; Jun, Y. C.; White, J. S.; Brongersma, M. L. Nat. Mater. 2010, 9, 193. (8) Lee, J. H.; You, M. H.; Kim, G. H.; Nam, J. M. Nano Lett. 2014, 14, 6217. (9) Li, J.; Cushing, S. K.; Meng, F.; Senty, T. R.; Bristow, A. D.; Wu, N. Nat. Photonics 2015, 9, 601. (10) Christopher, P.; Xin, H. L.; Linic, S. Nat. Chem. 2011, 3, 467. (11) Li, Z. T.; Zhu, Z. N.; Liu, W. J.; Zhou, Y. L.; Han, B.; Gao, Y.; Tang, Z. Y. J. Am. Chem. Soc. 2012, 134, 3322. (12) Linic, S.; Aslam, U.; Boerigter, C.; Morabito, M. Nat. Mater. 2015, 14, 567. (13) Brongersma, M. L.; Halas, N. J.; Nordlander, P. Nat. Nanotechnol. 2015, 10, 25. (14) Tian, Y.; Tatsuma, T. J. Am. Chem. Soc. 2005, 127, 7632. (15) Mubeen, S.; Lee, J.; Singh, N.; Kramer, S.; Stucky, G. D.; Moskovits, M. Nat. Nanotechnol. 2013, 8, 247. (16) Wen, F.; Zhang, Y.; Gottheim, S.; King, N. S.; Zhang, Y.; Nordlander, P.; Halas, N. J. ACS Nano 2015, 9, 6428. (17) Kwon, S. G.; Krylova, G.; Phillips, P. J.; Klie, R. F.; Chattopadhyay, S.; Shibata, T.; Bunel, E. E.; Liu, Y. Z.; Prakapenka, V. B.; Lee, B.; Shevchenko, E. V. Nat. Mater. 2015, 14, 215. (18) Jin, M.; He, G.; Zhang, H.; Zeng, J.; Xie, Z.; Xia, Y. Angew. Chem., Int. Ed. 2011, 50, 10560. (19) Liu, N.; Tang, M. L.; Hentschel, M.; Giessen, H.; Alivisatos, A. P. Nat. Mater. 2011, 10, 631. (20) Perez-Gonzalez, O.; Zabala, N.; Borisov, A. G.; Halas, N. J.; Nordlander, P.; Aizpurua, J. Nano Lett. 2010, 10, 3090. (21) Dong, S.; Zhang, K.; Yu, Z. P.; Fan, J. A. ACS Nano 2016, 10, 6716. 10183
DOI: 10.1021/jacs.7b04202 J. Am. Chem. Soc. 2017, 139, 10180−10183
