Geometry-Modulated Magnetoplasmonic Optical ... - ACS Publications

0 downloads 0 Views 448KB Size Report
Sep 27, 2017 - divided by ASPR/LSPR (absorbance of SPR peak for GNSs or LSPR peak for GNRs). Nano Letters. Letter. DOI: 10.1021/acs.nanolett.7b02583.
Letter pubs.acs.org/NanoLett

Geometry-Modulated Magnetoplasmonic Optical Activity of Au Nanorod-Based Nanostructures Bing Han,†,‡ Xiaoqing Gao,‡ Lin Shi,‡ Yonglong Zheng,‡ Ke Hou,‡ Jiawei Lv,‡ Jun Guo,‡ Wei Zhang,*,§ and Zhiyong Tang*,‡ †

College of Environmental Science and Engineering, North China Electric Power University, Beijing 102206, People’s Republic of China ‡ CAS Key Laboratory of Nanosystem and Hierarchical Fabrication, CAS Center for Excellence in Nanoscience, National Center for Nanoscience and Technology, Beijing 100190, People’s Republic of China § Institute of Applied Physics and Computational Mathematics, Beijing 100088, People’s Republic of China S Supporting Information *

ABSTRACT: Comprehension and modulation of optical activity at nanoscale have attracted tremendous interest in the past decades due to its potential application in many fields including chemical/biological sensing, artificial metamaterials, asymmetric catalysis, and so forth. As for the conventional molecular materials, magnetic field is among the most effective routes in inducing and manipulating their optical activity; whereas the magnetic optical activity at nanoscale calls for deeper understanding, especially for anisotropic noble metal nanoparticles. In this work, distinctly different magnetic circular dichroism (MCD) responses are demonstrated in gold nanorods (GNRs) with a derivative-shaped MCD signal corresponding to the transverse surface plasmon resonance (TSPR) band and a Gaussian-shaped signal at the position of the longitudinal surface plasmon resonance (LSPR) band. Furthermore, changing the aspect ratio of GNRs easily regulates such magnetoplasmonic CD response. More interestingly, GNR assemblies with different geometric configuration (end-to-end and side-by-side) show structure-sensitive magnetoplasmonic CD response. Armed with theoretical calculation, we clearly elucidate the intrinsic relationship of the resultant magnetoplasmonic CD response with the optical symmetry and geometry factor inside one-dimensional GNRs. This work not only greatly benefits our understanding toward the nature of SPR mode in anisotropic plasmonic nanostructures but also opens the way to achieve tunable magnetoplasmonic response, which will significantly advance the design and application of optical nanodevices. KEYWORDS: Au nanorods, self-assembly, geometry, magnetic circular dichroism, optical activity

P

Noteworthy, at the molecular scale MCD spectroscopy has shown the unique capability to reveal the degeneracy information on electronic energy levels, which cannot be derived from CD spectroscopy or light absorption spectroscopy.17−20 It is known that the degeneracy of excited states is directly affected by the symmetry of molecules, so MCD signals are highly sensitive to the structural change of molecules.13,19 By far, comprehension of molecular MCD is quite mature. However, knowledge of the origin of MCD response at nanoscale stays in the preliminary stage. Especially, the existent experimental design and theoretical study have mostly focused on the simple model with spherical NPs.21−23 Therefore, the first imperative question to be answered is whether MCD spectroscopy is effective in revealing symmetry in nanosystem. It has been widely studied that symmetry modulation in noble metal NPs gives rise to intricate plasmonic resonance behavior.24−26 Particularly, anisotropic noble metal NPs,

lasmonic circular dichroism (CD) response, the emerging optical activity at nanoscale, has become one of the hottest research topics in chemistry, physics, and materials due to its scientific and technological importance.1−4 Such optical activity is generated either by conjugation of noble metal nanoparticles (NPs) with chiral molecules or by constructing noble metal NP assemblies with chiral arrangement.5−10 Therefore, analogously to conventional molecular CD spectrum, plasmonic CD in noble metal nanostructures can be also classified into natural CD that requires chiral or helical distribution of electric charges in space.11,12 In contrast, there is an external and general way to generate CD activity in matters, which has been largely ignored in nanostructures. By applying a magnetic field parallel to the direction of light propagation, all matters regardless of with or without chirality exhibit CD effect under circularly polarized light.13,14 The universal optical activity is termed as magnetic circular dichroism (MCD), which originates from breaking of time-reversal symmetry by a magnetic field. On this basis, external magnetic field would be expected to provide a general and powerful route to achieve plasmonic optical activity at nanoscale.15,16 © 2017 American Chemical Society

Received: June 19, 2017 Revised: September 23, 2017 Published: September 27, 2017 6083

DOI: 10.1021/acs.nanolett.7b02583 Nano Lett. 2017, 17, 6083−6089

Letter

Nano Letters which possess multiple surface plasmon resonance (SPR) bands corresponding to electron oscillation along different geometrical directions,27−29 have exhibited fantastic magnetoplasmonic properties.30,31 Followed by the first report on the experimental MCD spectra of gold nanorods (GNRs),32 several pioneering works have verified that noble metal nanorods are excellent candidates to study the relationship between geometry/symmetry of NPs and the resultant magnetoplasmonic CD response.30,33 Furthermore, recent studies have convincingly indicated the importance of both the geometric anisotropy of ferromagnetic elliptical nanodisk and its periodic arrangement induced electromagnetic coupling in actively manipulating the magneto-optical activity (magneto-optical Kerr effect).34−36 Enlightened by these studies, another critical question is whether the magnetoplasmonic CD response in noble metal nanostructures can be modulated by manipulating geometry factor such as shape, dimension, and interparticle arrangement of constituent NPs. Unlike spherical NPs that are highly symmetric in both structure and interaction,21,23 anisotropic noble metal NPs like GNRs have the unique ability to assemble into a number of different aligned configurations, leading to intricate but interesting optical anisotropy.37−41 This configuration modulated optical property of GNR assemblies across the visible and near-infrared (NIR) spectral region makes them the ideal model to explore the key factors in manipulation of magnetoplasmonic CD, which will significantly benefit future development in active plasmonics. Hence, GNRs are selected in this work to thoroughly study the magnetoplasmonic CD effect at nanoscale. Impressively, distinct MCD response is discerned in GNRs for SPR modes along different directions. GNRs display a derivative-shaped MCD signal corresponding to the transverse SPR (TSPR) band, while a Gaussian-shaped signal appears at the position of the longitudinal SPR (LSPR) band. Both experimental result and theoretical simulation reveal that the geometry/symmetry dependent SPR coupling of TSPR and LSPR bands gives rise to the different magnetoplasmonic effect. Moreover, such magnetoplasmonic CD signal can be manipulated either by tuning the aspect ratio (AR) of GNRs or by fabricating GNR assemblies with different configurations, i.e. side-by-side (SS) and end-to-end (EE). The crucial factor in both modulation routes is explored to be the geometry factor in GNRs or GNR assemblies. GNRs were synthesized following the previously reported seed-mediated growth method,42 while spherical gold nanoparticles (GNSs) were also prepared as the contrast samples. As shown in Figure 1a,c, both as-synthesized GNSs and GNRs are well dispersed in aqueous solution and have narrow size distribution. The GNSs of 18.7 nm in diameter exhibit a sharp SPR adsorption band at around 520 nm (bottom curve in Figure 1b). Correspondingly, their magnetoplasmonic CD signal is a derivative shaped peak with its crossing zero point at the position of the SPR band (∼520 nm) (top curve in Figure 1b), originating from interaction of the circular plasmonic modes with an external magnetic field.21 In detailed analysis, the original degenerate circular plasmonic modes in GNSs split under the magnetic field along light propagation direction; as a result, GNSs adsorb left and right circularly polarized light (LCP and RCP) in different frequency, resulting in the derivative line shape. As comparison, GNRs with an AR of 3.0 (length and diameter are 46.7 nm 15.6 nm, respectively) show obviously different MCD response for their TSPR and LSPR

Figure 1. (a) TEM image and (b) MCD spectrum (top curve) and UV−vis-NIR spectrum (bottom curve) of well-dispersed GNSs. (c) TEM image and (d) MCD spectrum (top curve) and UV−vis-NIR adsorption spectrum (bottom curve) of well-dispersed GNRs with AR of 3.0; scale bar in (a,c) is 100 nm. Note that all the MCD spectra are plotted in term of ΔA (difference between absorbance of left and right circularly polarized light in the presence of a magnetic field) divided by ASPR/LSPR (absorbance of SPR peak for GNSs or LSPR peak for GNRs) and B (magnitude of the magnetic field), and MCD spectra are measured under the applied magnetic field of B = 1.6 T. All the UV− vis-NIR adsorption spectra are plotted in term of A (absorbance) divided by ASPR/LSPR (absorbance of SPR peak for GNSs or LSPR peak for GNRs).

bands. Top curve in Figure 1d manifests that GNRs also possess a derivative-shaped signal near the TSPR band, which is quite similar to the magnetoplasmonic CD signal of GNSs. However, at the position of LSPR band, the magnetoplasmonic CD response of GNRs is of Gaussian line shape. As further indicated in Figure S1, when the magnetic field is switched oppositely, MCD signal of GNRs is completely reversed while their UV−vis-NIR adsorption peak remains unchanged, confirming that the MCD signals result from real magnetooptical response rather than experimental artifact. Also, the intensity of MCD signal is linearly dependent on the magnitude of applied magnetic field, implying highly magnetic field sensitive magnetoplasmonic CD response in GNRs (Figure S2). It is noteworthy that such distinctive spectral behavior for TSPR and LSPR modes in GNRs have never been recognized in traditional spectroscopy such as UV−vis-NIR spectra, in which only two single peaks are discerned at different wavelengths due to the plasmon oscillation along either longitudinal or transverse axis. Therefore, the rich magnetoplasmonic CD response in GNRs system must reflect additional information related to the nature of SPR mode, which is stimulated by the external magnetic field. Such interesting finding in MCD signal pushes us to explore the intrinsic relationship between the magnetoplasmonic property and the geometry factor of GNRs. Then, GNRs of different AR from 2.1 to 3.9 are synthesized, and Figure S3 and Table S1 summarize the statistical size distribution of as prepared GNRs. It is clear that all the GNRs with varied AR are very uniform with narrow size distribution. The subsequent examination by MCD spectroscopy draws two important conclusions (Figure 2): (1) With the AR of GNRs increasing, 6084

DOI: 10.1021/acs.nanolett.7b02583 Nano Lett. 2017, 17, 6083−6089

Letter

Nano Letters

Figure 2. TEM images of GNRs with aspect ratio of (a) 2.1, (b) 2.4, (c) 3.0, (d) 3.7, (e) 3.9; scale bar = 100 nm, and corresponding MCD spectra (f) and UV−vis-NIR spectra (g) of well-dispersed GNRs with different aspect ratio. MCD data are measured under the applied magnetic field of B = 1.6 T.

difference between the coupled states. Accordingly, TSPR and LSPR modes of GNRs can be identified as the excited states in NPs, because they originate from oscillation of free electrons along the transverse or longitudinal axis. The different modes have different symmetry because of the anisotropic shape of GNRs. Seen from the left scheme in Figure 3a, GNRs possess

the bisignate magnetoplasmonic CD peaks corresponding to the TSPR band keep almost unchanged, whereas the Gaussianshaped MCD signals red shift with the LSPR bands (curve in Figure 2f). It is noticed that the shift of the magnetoplasmonic CD response shares the same AR-dependent change with the UV−vis-NIR adsorption peaks (curves in Figure 2g), which convincingly demonstrate that the magnetoplasmonic CD signal originates from interaction between the SPR modes and an external magnetic field. (2) When the AR of GNRs increases, the anisotropic factor (g factor) for either the transverse or longitudinal magnetoplasmonic CD signals decreases (Figure 2f and Table S3). The decreasing trend of the MCD intensity with AR increasing is quite different from that of the light absorption of GNRs, which is of enhanced extinction coefficient ε with AR increasing.43 Why does distinct magnetoplasmonic CD responses appear for different SPR modes of GNRs, and what is the reason causing decreased MCD intensity with AR increasing? Similar MCD line shape difference is discerned in triangular silver nanoprisms, which is attributed to the dipolar and quadrupolar SPR modes, respectviely.31 In contrast, both TSPR and LSPR modes of GNRs originate from dipolar resonance,44−46 which is totally different from those in Ag nanoprisms. Thus, there must exist an unexplored and major factor contributing to the distinct magnetoplasmonic CD signals at the position of TSPR and LSPR band for GNRs. Since MCD spectroscopy of molecules has been widely studied, the well-accepted principles may be adopted to give a preliminary explanation on the magnetoplasmonic CD response of GNRs. As for molecules, line shape of their MCD response is known to be determined by degeneracy of the excited or ground state. On one hand, if a molecule possesses a proper or improper rotation axis of the order greater than three, excited-state degeneracy is possible.13 The degenerate excited states will give rise to derivative-shaped MCD signal, which is recognized as “A term”.13,47 On the other hand, under magnetic field, the excited states might couple with each other, which results in Gaussian-shaped “B term”. In the case of low-symmetry molecules with rotation axis of the order two or one, there will be only nondegenerate excited states, so only “B terms” are expected in the MCD spectra.13,47 The intensity of “B term” is inversely proportional to the energy

Figure 3. Scheme of symmetry of (a) TSPR or LSPR mode in GNRs, and (b) proposed MCD response corresponding to TSPR (blue-filled area) and LSPR band (orange-filled area); the red curve in (b) represents the experimental MCD signal.

cylindrical symmetry for rotation along the longitudinal axis. As comparison, the rotational symmetry is lowered for rotation along the transverse axis, which is a 2-fold rotational axis (right scheme in Figure 3a). Therefore, the LSPR mode of a GNR is nondegenerate whereas its TSPR mode is doubly degenerate.48 LCP and RCP excite circular TSPR modes of GNRs, which are doubly degenerate in the absence of magnetic field. When an external magnetic field is applied along the direction of circularly polarized light propagation, the degeneration is lifted. As shown in Figure 3b, the adsorption of LCP and RCP for the two circular TSPR modes is shifted in energy, leading to a derivative-shaped signal near the position of the TSPR adsorption peak (blue-filled areas in Figure 3b). It is noted that the origin of the magnetoplasmonic CD signal for the 6085

DOI: 10.1021/acs.nanolett.7b02583 Nano Lett. 2017, 17, 6083−6089

Letter

Nano Letters

Figure 4. TEM images of SS GNR assembly (a) and EE GNR assembly (c); scale bar = 50 nm. CD spectra (top curves) and UV−vis-NIR spectra (bottom curves) of SS (red curves in (b)) or EE (pink curves in (d)) assembled GNRs with AR of 2.3. Blue curves stand for the corresponding spectra of dispersed GNRs, as reference. MCD data are measured under the applied magnetic field of B = 1.6 T.

shifts (top red curve in Figure 4b). In regard to EE GNR assembly, its LSPR peak displays the considerably reduced intensity and bathochromic shift from 672 to 762 nm (bottom pink curve in Figure 4d).49 Accordingly, the longitudinal MCD peak of EE GNR assembly shows the obviously reduced intensity and red shift to 770 nm (top pink curve in Figure 4d). To obtain full comprehension, GNRs with different AR are used as building blocks for assembly (Figure S4 and Table S2). As displayed in Figure S5−S8, for all the GNRs used, similar change tendency in MCD signals is distinguished. Very interestingly, the quantitative analysis on MCD spectra of GNR assemblies reveals that g factor of the GNR assembly has different change compared with the dispersed GNR. Because the strong MCD modulation appears for the LSPR mode, the longitudinal magnetoplasmonic CD response is selected for comparison. As summarized in Table S4 and S5, g factor of the longitudinal magnetoplasmonic CD signal is enhanced after GNRs form SS assembly, whereas it is reduced in EE GNR assembly. It can be concluded that for GNR assemblies, when the LSPR moves to lower energy, the magnetoplasmonic CD signal decreases; conversely, the MCD intensity increases. Evidently, the similar trend is found for dispersed GNRs, in which magnetoplasmonic CD response decreases along with red-shifted LSPR due to the increased AR. Altogether, one can clearly recognize the similarity between molecular MCD and magnetoplasmonic CD response. On one hand, taking the basic principle in MCD theory, the analysis on the MCD response of dispersed GNRs is reasonable. On the other hand, both dispersed GNRs and GNR assemblies have shown geometry or conformation-dependent magnetoplasmonic response. However, it must be pointed out that though molecular MCD and magnetoplasmonic CD are all based on symmetry principle, there are several different aspects between them. The traditional MCD theory deals with transition in

TSPR mode is quite similar to that of the GNSs, which exhibit rotational symmetry at any directions. But for the nondegenerate LSPR mode, the derivative-shaped signal will not be observed. Under magnetic field, the nondegenerate LSPR mode could couple with TSPR mode, generating a negative peak at the position of the LSPR mode and a positive peak at the position of TSPR mode (orange-filled areas in Figure 3d). The rationalization of regarding the interaction between TSPR mode and LSPR mode under magnetic field as B-like term has been verified in the recent work.30 Moreover, when the AR of GNRs increases, the energy difference between LSPR and TSPR becomes larger. Thus, the MCD signal arising from the magnetic field induced coupling will decrease in magnitude due to the weaker coupling between LSPR and TSPR. The fact that GNRs well dispersed in solution exhibit geometry-dependent magnetoplasmonic CD response encourages us to further investigate how the conformation of GNR assemblies affects the resultant MCD signal. With its anisotropic shape, dispersed GNRs are able to assemble into different patterns in solution with help of organic linkers, among which EE and SS are mostly studied. GNR assemblies have shown the conformation dependent optical property.30,49 Here we monitor the MCD response in EE and SS assembly. UV−vis-NIR absorption spectra (bottom curves in Figure 4b,d) and transmission electron microscopy (TEM) images (Figure 4a,c) confirm the successful formation of EE and SS GNR assemblies. For instance, as indicated by the bottom red curve in Figure 4b, compared with the UV−vis-NIR adsorption spectrum of dispersed GNRs (bottom blue curve in Figure 4b), the intensity of the LSPR adsorption decreases and blue shifts from 654 to 622 nm, which is characteristic of SS assembled GNRs.49 Correspondingly, the longitudinal MCD peak of SS GNR assembly blue shifts and its g factor increases compared with the dispersed GNRs, while the transverse MCD signal red 6086

DOI: 10.1021/acs.nanolett.7b02583 Nano Lett. 2017, 17, 6083−6089

Letter

Nano Letters

shape anisotropy gives rise to TSPR and LSPR modes at different resonant frequency. On one hand, corresponding to the TSPR adsorption peak, a derivative-shaped MCD signal is produced due to the double degeneracy. On the other hand, coupling between LSPR and TSPR mode is induced by external magnetic field, which leads to new MCD signals located at the frequency of LSPR and TSPR adsorption peak (part S3.1 in Supporting Information). To sum up, the MCD response of GNRs is a bisignate peak at the position of TSPR adsorption peak and a single peak located at the frequency of LSPR adsorption (Figures 1d and 3b). To understand the dependence of the optical property of GNRs on their geometry like AR is quite important. Figure 5

molecules, but surface plasmon in noble metal NPs cannot be considered as transition. What’s more, in GNR assemblies the existence of coupling between NRs makes the explanation on their magnetoptical activity more complicated, which is very difficult to be deduced from basic MCD theory. Therefore, we developed a theory to understand the magnetoplasmonic CD response of dispersed GNRs as well as how the conformation of GNR assemblies influence the resultant MCD effect. We first consider the case of the wavevector of circularly polarized light along the z-direction and the long axis of a GNR along x-direction (Figure S9). The MCD signal of GNRs comes from the magnetic field-induced difference of optical adsorption between LCP and RCP, which can be expressed as MCDGNR = Abs(LCP) − Abs(RCP)

(1)

where Abs = −εmωV Im[E α̂ E], ω is the frequency of the incident light,E = i1η with η = ±1 for LCP or RCP ̂ L̂ε ̂ + εm(I ̂ − L̂)I ]̂ −1 α̂ = V (ε ̂ − εmI )[ (2) +

( )

⎛ Lx 0 ⎞ ⎜ ⎟, L and Ly are the ⎝ 0 Ly ⎠ x geometry factor for SPR mode along the longitudinal and transverse direction, respectively. V is the volume of the GNR, ε and εm are the dielectric constants of Au and the medium, respectively. By combining eq 1 and eq 2 (more details in part S3.1 in Supporting Information), we obtain the following equations

In eq 2, I ̂ is the unit matrix, L̂ =

MCDGNR = Abs(η = 1) − Abs(η = − 1) = −

Figure 5. Calculated MCD spectra of dispersed GNRs with different AR.

presents theoretical simulation of MCD signals of GNRs with varied AR, which reproduces the line shape of the experimental data very well (Figure 2f). As the AR of GNRs becomes larger, the geometry factor Lx decreases

4εmωV f (ε) Im LxLy Δ0 (3)

( Lx =

where

(4)

here f (ω) =

(ωc =

eB m

is the cyclotron frequency).

Lx2 λ3 Ly − Lx

and e =

1 − 1/AR2 ). As

p

x

p

y

which is caused by the weaker coupling between the LSPR and TSPR modes under magnetic field. Impressively, the quantitatively good agreement of g factor (up to an overall constant) between experiment and theory confirms the credibility of our model (Table S3). In regard of GNR assemblies, the existence of coupling between GNRs must be considered (part S3.2 in Supporting Information). A pair of coupled GNRs is considered with dipole moment pi:

As for GNRs with large AR, the anisotropic factor at LSPR is g∝

(−1 + 21e ln 11 +− ee ),

demonstrated by the explicit dependence of the g factor on the geometry factor of GNRs in eq 5, the reduced geometry factor Lx correlates with the decreased g factor. This is also consistent with the fact that as for GNRs with larger AR, the increased difference between ωx and ωy leads to the decreased 1 longitudinal MCD (MCDl ∝ ω 2 / ω2 − ω 2 / ω2 ) (Figure S9b),

⎛ 1 − Ly ⎞ 1 − Lx ⎞ ⎛ Δ0 = ⎜εxx + εm⎟ ·⎜⎜εyy + εm⎟⎟ − f 2 (ω) Lx Ly ⎝ ⎠⎝ ⎠ ω A(ω) ω +c iγ

1 − e2 e2

(5)

Notably, based on our MCD theory one can see that the MCD signal is proportional to the magnetic field (in eq 3 and 4, ωc is proportional to B), and strongly depends on the geometry (Lx and Ly) of the structure. MCD peaks are determined by ReΔ0 = 0. It implies that MCD signal is attributed to the interplay of the SPR modes along different directions. For a sphere like GNS (ωx = ωy = ωz), the original degenerate SPR modes split under magnetic field and adsorb resonantly LCP or RCP at different frequency (ω+ or ω−, part S3.1 in Supporting Information) near the SPR mode. But for elongated particles like GNRs, the 3-fold degeneracy of these modes is lifted with doubly degenerate transverse SPR mode and nondegenerate longitudinal SPR mode. For this case, ωx ≠ ωy = ωz, MCD signals appear at both LSPR mode and TSPR mode of GNRs. As a result, for GNSs a derivative-shaped signal can be observed with its crossing zero corresponding to the SPR adsorption peak. However, with respect to GNRs, the

pi = αiE0(i = 1, 2)

(6)

where E0 is the incident field and αi is the polarizability. Taking into account the electromagnetic coupling between the dipoles, the dipole moment can be simplified as p = α eff E 0 q·V L 1 1 ε withα = Veff + V ε −mε , Leff = L + 3 , L is the geometry eff

1

m

R

factor of dispersed GNRs, and q is a constant determined by the assembly configuration. Therefore, the coupling between GNRs produces an effective geometry factor Leff. For the SS configuration, q = 1 and thus Leff > L, resulting in the blue shift of longitudinal MCD peak with increased g factor; for EE configuration, q = −2 and hence Leff < L, leading to the red shift 6087

DOI: 10.1021/acs.nanolett.7b02583 Nano Lett. 2017, 17, 6083−6089

Nano Letters of longitudinal MCD peak with decreased g factor. This analysis is very consistent with experimental observations (Figure 4b,d). In the more accurate theoretical model, the effect of multiple poles is taken into account by renormalization of the parameter q. By using fitting parameters for SS or EE configuration, excellent agreement between theory and experiment has been obtained, as demonstrated in the Tables S4 and S5. Overall, the conformation-dependent MCD response in GNR assembly is ascribed to the increase or decrease of the effective geometry factor in SS or EE configuration, respectively. This is coincident with the influence of geometry factor on the resultant MCD response in dispersed GNRs. In conclusion, geometry-modulated magnetoplasmonic CD response is explored in well-dispersed GNRs and GNR nanoassemblies. The anisotropic shape of GNRs leads to different symmetry between TSPR and LSPR modes, which gives rise to MCD response with distinct line shape. The direct relationship between symmetry of SPR mode and the spectral information is well established through MCD spectroscopy, which has never been recognized in previous UV−vis-NIR or CD spectra. The position and magnitude of such magnetoplasmonic CD response can be further manipulated by either tuning the AR of GNRs or fabricating GNR assemblies with different configurations. Both experimental and theoretical results disclose that the geometry factor in GNRs or nanoassemblies is the crucial factor to produce highly tunable magnetoplasmonic CD signals. Our work highlights that MCD spectroscopy will be a powerful method to discriminate SPR modes with different symmetry in anisotropic metal nanostructures. Furthermore, understanding the importance of geometry factors lays the foundation to achieve sophisticated modulation of the magnetoplasmonic response at nanoscale, which will greatly benefit their applications in nanophotonic devices and refractometric sensing.





ACKNOWLEDGMENTS



REFERENCES

This work was supported by National Key Basic Research Program of China (Grants 2014CB931801, Z.Y.T.; 2013CB632805, W.Z.), National Key Research and Development Program of China (Grants 2016YFA0200700, Z.Y.T.; 2017YFA0303400, W.Z), National Natural Science Foundation of China (Grants 21475029, 91427302, Z.Y.T.; 11174042, 11374039 and 11774036, W.Z.; 21703064, B.H.), Science Fund for Creative Research Groups of the National Natural Science Foundation of China (21721002, Z.Y.T.), Frontier Science Key Project of the Chinese Academy of Sciences (Grant QYZDJSSW-SLH038, Z.Y.T.), Instrument Developing Project of the Chinese Academy of Sciences (Grant YZ201311, Z.Y.T.), CASCSIRO Cooperative Research Program (Grant GJHZ1503, Z.Y.T.), the “Strategic Priority Research Program” of Chinese Academy of Sciences (Grant XDA09040100, Z.Y.T.), K. C. Wong Education Foundation (Z.Y.T.) and the Fundamental Research Funds for the Central Universities (Grant 2017MS046, B.H.).

(1) Gansel, J. K.; Thiel, M.; Rill, M. S.; Decker, M.; Bade, K.; Saile, V.; Von Freymann, G.; Linden, S.; Wegener, M. Science 2009, 325, 1513. (2) Kuzyk, A.; Schreiber, R.; Fan, Z. Y.; Pardatscher, G.; Roller, E. M.; Hogele, A.; Simmel, F. C.; Govorov, A. O.; Liedl, T. Nature 2012, 483, 311. (3) Guerrero-Martínez, A.; Alonso-Gómez, J. L.; Auguié, B.; Cid, M. M.; Liz-Marzán, L. M. Nano Today 2011, 6, 381. (4) Ben-Moshe, A.; Maoz, B.; Govorov, A. O.; Markovich, G. Chem. Soc. Rev. 2013, 42, 7028. (5) Mastroianni, A. J.; Claridge, S. A.; Alivisatos, A. P. J. Am. Chem. Soc. 2009, 131, 8455. (6) Bao, Z. Y.; Zhang, W.; Zhang, Y. L.; He, J.; Dai, J.; Yeung, C. T.; Law, G. L.; Lei, D. Y. Angew. Chem. 2017, 129, 1303. (7) Kuzyk, A.; Schreiber, R.; Zhang, H.; Govorov, A. O.; Liedl, T.; Liu, N. Nat. Mater. 2014, 13, 862. (8) Lan, X.; Lu, X.; Shen, C.; Ke, Y.; Ni, W.; Wang, Q. J. Am. Chem. Soc. 2015, 137, 457. (9) 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. (10) Ma, W.; Kuang, H.; Xu, L.; Ding, L.; Xu, C.; Wang, L.; Kotov, N. A. Nat. Commun. 2013, 4, 2689. (11) Fan, Z.; Govorov, A. O. Nano Lett. 2010, 10, 2580. (12) Govorov, A. O.; Fan, Z. Y.; Hernandez, P.; Slocik, J. M.; Naik, R. R. Nano Lett. 2010, 10, 1374. (13) Mason, W. R. A Practical Guide to Magnetic Circular Dichroism Spectroscopy; Wiley-Interscience: New York, 2007. (14) Stephens, P. J. J. Chem. Phys. 1970, 52, 3489. (15) Armelles, G.; Caballero, B.; Prieto, P.; Garcia, F.; Cebollada, A.; Gonzalez, M. U.; Garcia-Martin, A. Nanoscale 2014, 6, 3737. (16) Armelles, G.; Cebollada, A.; García-Martín, A.; González, M. U. Adv. Opt. Mater. 2013, 1, 10. (17) Bradley, J. M. J. Am. Chem. Soc. 2011, 133, 19676. (18) Kobayashi, N.; Nakai, K. Chem. Commun. 2007, 40, 4077. (19) Muranaka, A.; Yokoyama, M.; Matsumoto, Y.; Uchiyama, M.; Tsuda, A.; Osuka, A.; Kobayashi, N. ChemPhysChem 2005, 6, 171. (20) Ziegler, C. J.; Erickson, N. R.; Dahlby, M. R.; Nemykin, V. N. J. Phys. Chem. A 2013, 117, 11499. (21) Pineider, F.; Campo, G.; Bonanni, V.; Fernandez, C. J.; Mattei, G.; Caneschi, A.; Gatteschi, D.; Sangregorio, C. Nano Lett. 2013, 13, 4785. (22) Shiratsu, T.; Yao, H. J. Nanophotonics 2016, 10, 046004. (23) Yao, H.; Shiratsu, T. Nanoscale 2016, 8, 11264. (24) Prodan, E.; Radloff, C.; Halas, N. J.; Nordlander, P. Science 2003, 302, 419.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.7b02583. Materials and methods: synthesis of GNRs and controlled assembly of GNRs into SS mode and EE mode; experimental data of GNRs or GNR assemblies with different AR; theoretical calculation details of MCD theory for well-dispersed Au nanorods (GNRs) and GNR assemblies; additional figures and tables (PDF)



Letter

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. ORCID

Zhiyong Tang: 0000-0003-0610-0064 Author Contributions

All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest. 6088

DOI: 10.1021/acs.nanolett.7b02583 Nano Lett. 2017, 17, 6083−6089

Letter

Nano Letters (25) Lassiter, J. B.; Sobhani, H.; Fan, J. A.; Kundu, J.; Capasso, F.; Nordlander, P.; Halas, N. J. Nano Lett. 2010, 10, 3184. (26) Wang, H.; Brandl, D. W.; Nordlander, P.; Halas, N. J. Acc. Chem. Res. 2007, 40, 53. (27) Link, S.; El-Sayed, M. A. J. Phys. Chem. B 1999, 103, 8410. (28) Sherry, L. J.; Jin, R. C.; Mirkin, C. A.; Schatz, G. C.; Van Duyne, R. P. Nano Lett. 2006, 6, 2060. (29) Sosa, I. O.; Noguez, C.; Barrera, R. G. J. Phys. Chem. B 2003, 107, 6269. (30) Melnikau, D.; Govyadinov, A. A.; Sánchez-Iglesias, A.; Grzelczak, M.; Liz-Marzán, L. M.; Rakovich, Y. P. Nano Lett. 2017, 17, 1808. (31) Yao, H.; Shiratsu, T. J. Phys. Chem. C 2017, 121, 761. (32) Artemyev, M.; Krutokhvostov, R.; Melnikau, D.; Oleinikov, V.; Sukhanova, A.; Nabiev, I. Proc. SPIE 2012, 845729. (33) Armelles, G.; Cebollada, A.; García, F.; García-Martín, A.; de Sousa, N. ACS Photonics 2016, 3, 2427. (34) Maccaferri, N.; Bergamini, L.; Pancaldi, M.; Schmidt, M. K.; Kataja, M.; Dijken, S. v.; Zabala, N.; Aizpurua, J.; Vavassori, P. Nano Lett. 2016, 16, 2533. (35) Maccaferri, N.; González-Díaz, J. B.; Bonetti, S.; Berger, A.; Kataja, M.; van Dijken, S.; Nogués, J.; Bonanni, V.; Pirzadeh, Z.; Dmitriev, A.; Åkerman, J.; Vavassori, P. Opt. Express 2013, 21, 9875. (36) Lodewijks, K.; Maccaferri, N.; Pakizeh, T.; Dumas, R. K.; Zubritskaya, I.; Åkerman, J.; Vavassori, P.; Dmitriev, A. Nano Lett. 2014, 14, 7207. (37) Funston, A. M.; Novo, C.; Davis, T. J.; Mulvaney, P. Nano Lett. 2009, 9, 1651. (38) Han, B.; Zhu, Z. N.; Li, Z. T.; Zhang, W.; Tang, Z. Y. J. Am. Chem. Soc. 2014, 136, 16104. (39) Lee, A.; Ahmed, A.; dos Santos, D. P.; Coombs, N.; Park, J. I.; Gordon, R.; Brolo, A. G.; Kumacheva, E. J. Phys. Chem. C 2012, 116, 5538. (40) Lee, A.; Andrade, G. F. S.; Ahmed, A.; Souza, M. L.; Coombs, N.; Tumarkin, E.; Liu, K.; Gordon, R.; Brolo, A. G.; Kumacheva, E. J. Am. Chem. Soc. 2011, 133, 7563. (41) Zhu, Z.; Liu, W.; Li, Z.; Han, B.; Zhou, Y.; Gao, Y.; Tang, Z. ACS Nano 2012, 6, 2326. (42) Nikoobakht, B.; El-Sayed, M. A. Chem. Mater. 2003, 15, 1957. (43) Link, S.; Mohamed, M. B.; El-Sayed, M. A. J. Phys. Chem. B 1999, 103, 3073. (44) Chen, H.; Shao, L.; Li, Q.; Wang, J. Chem. Soc. Rev. 2013, 42, 2679. (45) Muskens, O. L.; Bachelier, G.; Fatti, N. D.; Vallée, F.; Brioude, A.; Jiang, X.; Pileni, M.-P. J. Phys. Chem. C 2008, 112, 8917. (46) Zuloaga, J.; Prodan, E.; Nordlander, P. ACS Nano 2010, 4, 5269. (47) Mack, J.; Stillman, M. J.; Kobayashi, N. Coord. Chem. Rev. 2007, 251, 429. (48) Tserkezis, C.; Papanikolaou, N.; Almpanis, E.; Stefanou, N. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 80, 125124. (49) Jain, P. K.; Eustis, S.; El-Sayed, M. A. J. Phys. Chem. B 2006, 110, 18243.

6089

DOI: 10.1021/acs.nanolett.7b02583 Nano Lett. 2017, 17, 6083−6089