Open Access
Add to BibSonomyAbstract
Arrays of differently sized disk shaped gold nanoantennas are prepared on glass, which show localized surface plasmon resonance and Rayleigh anomalies in the near infrared and telecom range between 1000 and 1500nm wavelength. The spectral position of these grating resonances depends critically on the period of the array and the size of the nanoantennas. When PbS quantum dots embedded in PMMA surround the nanoantennas, an up to four fold enhancement of the photoluminescence is observed at the grating resonances due to the constructive diffractive feedback among neighboring antennas. In accordance with the grating resonances a shift of the emission towards smaller wavelengths with decreasing disk diameter is demonstrated.
© 2016 Optical Society of America
1. Introduction and motivation
In the field of nonlinear optics, the possibility to enhance absorption or emission in a very small volume has been greatly influenced by high optical confinement and local field enhancement of localized surface plasmons resonances (LSPR) occurring at metal nanoparticles, nanostructures or nanoantennas [1–3].
Several theoretical [4,5] and experimental [4–9] studies have successfully shown array resonances of linear (1D) and areal (2D) arrangements of nanoantennas of generic plasmonic metals along with their extinction efficiencies. Furthermore the increase in emission and even lasing in the visible in very near IR (NIR) spectral range below 1100nm wavelength were observed in connection with these array resonances [10].
In this paper we explore the possibility of tailoring these characteristic array resonances towards longer wavelengths (telecom range). For this a systematic investigation for the effect of multiple scattering among regularly arranged metal nanoantennas along with the resulting enhancement of light emission from quantum dots placed in their vicinity has been carried out. Based on these results applications employing the enhanced emission due to nanoantenna arrays for the use in integrated silicon photonics could be envisaged.
2. Experimental
Arrays of gold nanoantennas with overall sizes of 300 μm x 300 μm were fabricated on a 1 mm thick glass substrate using electron beam lithography followed by thermal evaporation and a lift off process. Further fabrication details are presented in Appendix I. To improve the adhesion of the 50 nm Au layer a 2 – 3 nm thin Cr film was deposited first onto the glass as an interlayer. However this thin Cr-layer does not affect the localized surface plasmon resonance noticeably [6].
The samples considered in this paper contain circular nanoantennas with varying disk diameters (150 nm to 250 nm), which are arranged in a quadratic lattice with a period of 900 nm. Table 1 contains an overview of the sample parameters. The SEM-images in Fig. 1 show details of two (ø175 nm and ø 250 nm) of these disk shaped Au nanoantenna arrays revealing the perfect placement and high uniformity of the fabricated nanodisc antennas.
Table 1. Overview of the fabricated samples
Fig. 1 SEM Images of disk shaped gold nanoantennas on glass with (a) 175 nm and (b) 250 nm diameter. The nanoantennas are arranged in square lattices with a period of a = 900 nm.
To avoid smearing of optical spectral details, a homogeneous index environment with n = 1.5 for the nanoantennas is necessary [11–13]. This was achieved by covering and embedding the nanoantennas with a 3 μm ± 0.4 μm thick PMMA layer. For this a drop of 5 wt% PMMA chloroform solution was spin coated at 500 rpm for 1 min over the antenna arrays and left to dry overnight.
3. Nanoantenna transmission
After fabrication, the optical extinction of the nanoantennas was investigated by transmission measurements. A collimated beam from halogen lamp is used to illuminate the sample, the transmitted light is collected by Mitutoyo (infinity corrected long working distance) objective and the area of 300 μm x 300 μm of nanoantenna arrays is imaged on the fiber bundle with help of a lens system. The light from the fiber bundle is input into an Acton Spectra Pro SP-2500 Monochromator from Princeton Instruments and the data analyzed. First, nanoantennas with 250 nm diameter are considered. The experimentally determined transmission intensity is plotted in Fig. 2 as a black curve.
Fig. 2 Transmission spectra of nanoantenna disks with diameter 250 nm. The experimental curve is shown in black. (a) A Lorentzian fit (green) reveals the underlying single particle surface Plasmon resonance. (b) The theoretical COMSOL calculation for the nanoantenna/disk array (red) reveals the surface lattice resonance in excellent spectral correspondence with the experiment. The vertical blue dotted lines indicate the diffraction edges.
A broad extinction band is centered around 1040 nm and can be assigned to the excitation of the localized surface plasmon resonance (LSPR) of the single Au disks. Fitting a Lorentzian to this broad dip a full width at half maximum (FWHM) of approx. 570 nm is obtained [Fig. 2(a)]. This corresponds rather well to the spectral behavior of the theoretically calculated extinction cross section for a single gold disk (see Appendix II), which exhibits an extinction maximum at a wavelength of 1100 nm and yielding a FWHM of 610 nm. The calculated resonance spectrum shows clearly, that higher order resonances of a single particle (e.g. quadrapole etc.) do not play a role in the NIR. The dipole resonance clearly governs the overall extinction with scattering being the dominating contribution to the extinction. Comparing the level of the calculated extinction cross section with the transmission measurements also shows a good correspondence: Based on the calculated extinction cross section in Appendix II, a minimum transmission of 65% is expected while the experiment shows a minimum level of about 67% [Fig. 2(a)]. So, the broad resonance dips in the spectra can clearly be identified as the single disk resonances.
To study the collective effect of the periodic nanoantenna arrangement the diffraction edges of the 2D array have to be considered. These are indicated by the dashed vertical lines [Fig. 2(a)] at 955 nm and 1350 nm and indicate the maximum wavelengths for which new propagating diffraction orders appear. The first diffraction edge is situated at λ0 = n*a = 1.5 * 900 nm = 1350 nm and corresponds to the diffraction from the (1, 0) lattice planes [15–17], while the second diffraction edge occurs at 1.5 * 900 * sin (45°) = 955 nm and corresponds to the (1, 1) lattice planes. As these disks are arranged in a quadratic lattice, two (1, 0) and (1, 1) diffraction orders appears. Close to these diffraction edges specific features (anomalies) appear in the spectrum which leads to sharp change in the transmission. Especially at the long wavelength side of the (1, 0) diffraction edge at λ = 1400 nm a sharp transmission dip (extinction peak) is observed. Its fano-like line shape arises from the coupling of the single particle LSPRs within the periodic nanoantenna array. The sharp dip therefore represents a surface lattice resonance (SLR) and appears slightly red shifted from the diffraction edge [8,9,18]. This sharp extinction peak is also confirmed by COMSOL-calculations [red line in Fig. 2(b)], which were performed for the same arrangement of gold nanoantennas embedded in a 1.5 refractive index environment. In these calculations, the dips appear even sharper and more pronounced than in the experiment. This is partly due to infinitely extended array of nanoantennas assumed in the calculation using periodic boundary conditions. Additionally, in theory a perfect plane wave is incident upon the nanoantenna array, while under real experimental conditions we have a finite nanoantenna array and the collimation of the beam might not be perfect.
In contrast to the (1, 0) resonance, for the (1, 1) plane such a sharp lattice resonance cannot be observed. However the spectrum shows also some anomaly around the (1, 1) diffraction edge at λ = 955 nm and a spectral range of increased transmission surrounding this diffraction edge and disturbing the simple broad Lorentzian extinction of the single disk resonance in this spectral range.
The embedding of the nanoantennas in a homogeneous index environment is important for the appearance of the sharp lattice resonance and in [12, 13] it was demonstrated that with an increasing index mismatch between glass substrate and embedding (cover) material, the lattice resonance will soon disappear. A different refractive index of substrate and cladding leads to different wavelengths of light travelling through the substrate and cladding (between neighbouring antennas), thus preventing the dipolar scattering fields from interfering coherently.
Furthermore, a collimated irradiation of the nanoantenna arrays is crucial to observe lattice resonances in transmission, since only a plane wave excitation at normal incidence guarantees the excitation of all nanoantenna oscillations with the same phase. This is necessary to excite the collective surface lattice resonance of several hundred nanoantennas.
4. Impact of nanoantenna size
To investigate the impact of the antenna size on the surface lattice resonances (SLRs), Au-disk nanoantennas with diameters of 250 nm, 225 nm, 200 nm, 175 nm and 150 nm were investigated. The stacked transmission data with specific focus on the IR range are shown in Fig. 3(a).
Fig. 3 Experimental data from (a) Transmission experiments of nanoantennas embedded in PMMA (b) detail of the IR-transmission (c) micro-PL data of nanoantennas in vicinity of PbS quantum dots, in the range from 1300 nm to 1600 nm.
Again the single particle resonance is indicated by the broad overlaid Lorentzians which are centred at approximately 1050 nm, 1000 nm, 950 nm, 900 nm and 800 nm, respectively [Fig. 3(a), red crosses]. Relatively sharp (1, 0) surface lattice resonances (marked as blue crosses) are always observed at the long wavelength side of the broad resonance. These occur at approx. 1400 nm for 250 nm diameter disks and shift slightly to the blue as the diameter of the disks is reduced. It is also observed that, these sharp SLRs eventually decay strongly by the time, the disk diameter reaches 175 nm and are almost absent for 150 nm disks.
All this is connected with the blue-shift of the broad single particle resonance due to the decreasing nanoantenna size. To understand the slight blue shift of the SLR it is sufficient to study the impact of the first order scattering from nearest neighbour antennas: Considering only the 2 nearest neighbour antennas besides our specific nanoantenna, the dipole moment of the specific nanontenna can be written as
Here, ‘ϕ’ is the phase delay of the scattered field E1 with respect to the exciting field E0 due to the particle resonance and ‘a’ the period of the square lattice (distance between nearest neighbouring antennas). The phase shift ϕ, changes from nearly 0 for wavelengths much longer than the single particle plasmon, to π/2 at the single particle resonance and π for much shorter wavelengths. The SLR appears when the direct incident field E0 and the scattered field E1 from the neighbouring antennas interfere constructively:The further blue shifted the single particle resonance is from the diffraction edge, the smaller is the phase delay ϕ. This requires an increase of the term ‘ka’ to fulfil Eq. (2), which can only be achieved by a rising ‘k’ corresponding to a shorter wavelength of the SLR.The observed decay of the SLR with decreasing disk diameter is a direct consequence of the reduced scattering intensity of the single disks at the SLR wavelength (which is tied to the (1, 0) diffraction edge around λ = 1400 nm). For the 250 nm discs in Fig. 4(a) the scattering cross section still amounts to 1.2x105 nm2 at λ = 1400 nm it has decayed to 4.5x103 nm2, while for the 150 nm disks at the same wavelength [Fig. 4(b)]. There are two reasons for this drastic decay in scattering cross section at λ = 1400 nm: First, the smaller diameter of the disks leads to an overall reduced scattering cross section of the single disks, since there is less scattering volume (metal). From Fig. 4 it can be observed that the reduction in diameter from 250 nm to 150 nm already halves the peak scattering cross section. Second, the single disk resonance shifts to the blue with decreasing diameter enlarging the spectral distance between single disk resonance (LSPR) and SLR. The red markers in Fig. 3(a) clearly show the blue-shift of the single disk resonance for decreasing disk diameters down to a wavelength of 800 nm for the 150 nm disks. This corresponds very well to theoretical calculations in Fig. 4, which predict the peak extinction at λ = 810 nm for a disk with 150 nm diameter. Since the scattered intensity decays away from the single disk resonance the increased spectral distance between LSPR and SLR reduces the scattering at the SLR-wavelength further resulting in the overall strong decay of the SLR-feature with reduced disk size.
Fig. 4 Extinction (black), absorption (red) and scattering (blue) cross sections of Au-with with (a) 250 nm and (b) 150 nm diameter (based on ε (ω) from Johnson and Christy Data [19]). Clearly the dipole resonance dominates the optical response of a single nanoantenna in the NIR.
Now, let us consider the (1, 1) Raleigh anomaly (RA), which occurs at approx. 955 nm due to the grating effect. As the disk diameter decreases, the broad Lorentzian peak, shifts from 1050 nm to 800 nm and passes through the (1, 1) RA. This results in an interference of the two spectral features (Lorentzian and RA), leading to less pronounced transmission dips for the smaller nanoantennas.
5. Micro-photoluminescence (μ-PL)
Metal nanoantennas supporting LSPRs show a high degree of optical field confinement in the near field resulting in small mode volumes (V). Furthermore the sharp SLRs indicate a reduction of scattering losses and a considerable overall increase in the quality (Q)-factor of the collective resonance. The resulting increase of the Q/V-ratio should therefore lead to a sizeable Purcell-effect. Here we therefore investigate, if the luminescence of an emitter is enhanced, when its luminescence frequency coincides with the SLR.
As suitable emitter for the spectral range around 1400 nm we choose PbS quantum dots, received from Evident Technologies. Due to their size dispersion they show a photoluminescence in the range from 1000 nm to 1600 nm. (The PL spectra for PbS embedded in PMMA on glass are shown in Appendix III for reference.)
Various samples with different concentrations of PbS in PMMA were prepared and their photoluminescence (PL) is measured. The highest PL intensity is observed for the sample prepared by mixing 20 μlts of PbS-containing solution to 1 mlts of 5 wt% PMMA in chloroform giving 2*104 PbS nanoparticles per μm3 of PMMA. This solution is spin coated on the nanoantenna arrays with Au-disc diameters of 250 nm, 225 nm, 200 nm and 175 nm, which are again arranged in a square lattice with a = 900 nm. Using a spin speed of 500 rpm for 1 min a coating thickness of 3.2 ± 0.2 μm was achieved. The samples were left to dry overnight in Ar atmosphere.
First the transmission spectra of the PbS/PMMA-covered samples were taken, revealing the same spectral characteristics as already mentioned [Fig. 3(a), Fig. 3(b)]. Subsequently the PL-spectra of the nanoantenna samples with embedded PbS-particles were measured using a He-Ne laser (633nm). The resulting μ-PL spectra [Fig. 3(c)] show rather broad enhanced emission peaks on top of the wide unperturbed PbS-particle emission. Comparing these emission spectra with the transmission spectra from Fig. 3(b), a clear spectral correspondence of transmission dips (i.e. extinction peaks) and emission peaks is observed. Furthermore a similar trend of enhancement is observed: The strongest extinction resonance for the 250 nm nanoantennas also yields the strongest (nearly 4-fold) emission enhancement, while for the smallest particles extinction and emission resonances are not anymore observable. In addition, the slight blue shift of the SLRs with decreasing disc diameter can also be observed as blue shift of the emission peaks in the PL-spectra. This clearly shows that a luminescence enhancement due to the SLR is detected in the near IR within the range of the telecom wavelengths. To clearly prove if the emission enhancement is really based on an increase in radiative recombination rate due to the Purcell-effect or if an upwards re-direction of emission is the cause of the observed enhanced luminescence, time resolved PL measurements are necessary. This remains a challenging task for the future, since sensitive and fast time resolved measurements for wavelengths above 1000 nm are not standard.
We also carried out intensity dependent PL measurements by varying the incident power of a Femtosecond laser (ca. 120fs pulses) and observed a linear increase in luminescence intensity with rising pump power. Although the pump power was increased up to the damage threshold of the sample we could not observe a line narrowing or even lasing as reported by Zhou et al. [10] for the laser dye IR-140 around 900 nm wavelength. We suppose that the lower quantum efficiency of our PbS-particles and an overall small number of PbS-particles in the vicinity of the nanoantennas, which exhibit the necessary spectral overlap of emission wavelength and SLR, might be the reason for this.
6. Conclusion
In conclusion the coupling of gold nanoantenna resonances in the IR telecom range was studied. Different experimental sets with disk shaped nanoantennas of varying size were analysed. In all the samples, a broad single particle resonance and a sharp fano line shaped resonance are clearly visible. The fano shaped surface lattice resonances (SLR) were formed due to far field coupling of the single particle localized surface plasmon resonances (LSPR) within the square nanoantenna lattice and appeared close to the (1,0) diffraction edge. Reducing the size of the Au-disc lead to a blue shift of LSPR and SLR and a reduction in overall extinction cross section as well as in a decline of the SLR-feature. Surrounding the nanoantenna arrays with PbS-particles embedded in PMMA lead to the observation of an up to four fold enhanced luminescence. The observed joint shift of SLR-related transmission dip and luminescence peak proved their correlation and showed the possibility to enhance the luminescence at telecom wavelengths using easily tuneable plasmonic surface lattice resonances.
Appendix I
Sample fabrication pre and post e-beam lithography:
The glass is cleaned in an ultrasonic bath with acetone for 2 mins and then in isopropanol for 2 mins. A first layer, 600k PMMA (AR-P 669.04 from ALLRESIST GmbH) is spin coated at 5000 rpm for 90 s. Baking of this layer is carried out for 1 hr at 200 °C. The sample is then cooled down to room temperature and a second layer of 950k PMMA (AR-P 679.03 from ALLRESIST GmbH) is spin coated with the same parameters as before and baked again at 180 °C for 1 hr. The sample is cooled down to room temperature. After this, a conductive layer of SX AR−PC 5000/90.2 is deposited to avoid charging during e-beam lithography (spin coated at 2500 rpm for 90 s). Finally, the sample is baked at 105 oC for 5 mins.
The disc shaped structures are defined by electron-beam lithography using a RAITH PIONEER system. The structures are written with an acceleration voltage of 30 kV, a beam current of 150 pA and an area dose between 1000 and 1500 μC/cm2 depending on the structure sizes.
Nanoantennas of the same geometry were produced each in an area of 300 μm x 300 μm. After removing the conducting layer with water and developing for 1 min in isopropanol; 50 nm Au layer with 2–3 nm Cr as interlayer are deposited in a thermal evaporator onto the sample. Finally, the remaining photoresist is removed by acetone.
Appendix II
Using the finite-element software COMSOL for the simulation of plasmonic nanoparticles the absorption, scattering and overall extinction cross sections were calculated for two single gold disks with diameters 250 nm and 150 nm and a thickness of 50 nm.
Appendix III
Photoluminescence spectrum of PbS-particles dispersed in PMMA layer (without nanoanetennas). Due to the size distribution of the particles an in homogeneously broadened emission centered around 1420nm is observed (see Fig. 5).
Acknowledgments
The authors thank the International Max Planck Research School for Science and Technology of Nanostructures and the Federal Ministry for Education and Research (BMBF) for their financial support under project number FKZ:03Z2HN12, within the Centre for Innovation Competence SiLi-nano®.
References and links
1. V. Gavrilenko, Optics of Nanomaterials (Pan Stanford, 2010).
2. S. A. Maier, Plasmonics: Fundamentals and Applications (Springer, 2007).
3. S. A. Maier, P. G. Kik, H. A. Atwater, S. Meltzer, E. Harel, B. E. Koel, and A. A. Requicha, “Local detection of electromagnetic energy transport below the diffraction limit in metal nanoparticle plasmon waveguides,” Nat. Mater. 2(4), 229–232 (2003). [CrossRef] [PubMed]
4. S. Zou, N. Janel, and G. C. Schatz, “Silver nanoparticle array structures that produce remarkably narrow plasmon lineshapes,” J. Chem. Phys. 120(23), 10871–10875 (2004). [CrossRef] [PubMed]
5. A. B. Evlyukhin, C. Reinhardt, U. Zywietz, and B. N. Chichkov, “Collective resonances in metal nanoparticle arrays with dipole-quadrupole interactions,” Phys. Rev. B 85(24), 245411 (2012). [CrossRef]
6. B. Auguié and W. L. Barnes, “Collective Resonances in Gold Nanoparticle Arrays,” Phys. Rev. Lett. 101(14), 143902 (2008). [CrossRef] [PubMed]
7. S. Zou and G. C. Schatz, “Narrow plasmonic/photonic extinction and scattering line shapes for one and two dimensional silver nanoparticle arrays,” J. Chem. Phys. 121(24), 12606–12612 (2004). [CrossRef] [PubMed]
8. A. I. Kuznetsov, A. B. Evlyukhin, M. R. Gonçalves, C. Reinhardt, A. Koroleva, M. L. Arnedillo, R. Kiyan, O. Marti, and B. N. Chichkov, “Laser Fabrication of Large-Scale Nanoparticle Arrays for sensing Applications,” ACS Nano 5(6), 4843–4849 (2011). [CrossRef] [PubMed]
9. I. Aristov, U. Zywietz, A. B. Evlyukhin, C. Reinhardt, B. N. Chichkov, and A. V. Kabashin, “Laser-ablative engineering of phase singularities in plasmonic metamaterial arrays for biosensing applications,” Appl. Phys. Lett. 104(7), 071101 (2014). [CrossRef]
10. W. Zhou, M. Dridi, J. Y. Suh, C. H. Kim, D. T. Co, M. R. Wasielewski, G. C. Schatz, and T. W. Odom, “Lasing action in strongly coupled plasmonic nanocavity arrays,” Nat. Nanotechnol. 8(7), 506–511 (2013). [CrossRef] [PubMed]
11. B. Auguié, X. M. Bendaña, W. L. Barnes, and F. J. García de Abajo, “Diffractive arrays of gold nanoparticles near an interface: Critical role of the substrate,” Phys. Rev. B 82(15), 155447 (2010). [CrossRef]
12. S. V. Zhukovsky, V. E. Babicheva, A. V. Uskov, I. E. Protsenko, and A. V. Lavrinenko, “Electron photoemission in plasmonic nanoparticle arrays: analysis of collective resonances and embedding effects,” Appl. Phys., A Mater. Sci. Process. 11, 6929–6940 (2014).
13. S. V. Zhukovsky, V. E. Babicheva, A. V. Uskov, I. E. Protsenko, and A. V. Lavrinenko, “Enhanced Electron Photoemission by Collective Lattice Resonances in Plasmonic Nanoparticle-Array Photodetectors and Solar Cells,” Plasmonics 9(2), 283–289 (2014). [CrossRef]
14. U. Hohenester and A. Trugler, “MNPBEM: A Matlab toolbox for the simulation of plasmonic nanoparticles,” Comput. Phys. Commun. 183(2), 370–381 (2012). [CrossRef]
15. L. Rayleigh, “On the dynamical theory of grating,” Proc. R. Soc. Lond., A Contain. Pap. Math. Phys. Character 79(532), 399–416 (1907). [CrossRef]
16. G. Vecchi, V. Giannini, and J. Gómez Rivas, “Shaping the fluorescent emission by lattice resonances in plasmonic crystals of nanoantennas,” Phys. Rev. Lett. 102(14), 146807 (2009). [CrossRef] [PubMed]
17. G. Vecchi, V. Giannini, and J. G. Rivas, “Surface modes in plasmonic crytals induced by diffractive coupling of nanoantennas,” Phys. Rev. B 80(20), 201401 (2009). [CrossRef]
18. S. Murai, M. A. Verschuuren, G. Lozano, G. Pirruccio, S. R. K. Rodriguez, and J. G. Rivas, “Hybrid plasmonic-photonic modes in diffractive arrays of nanoparticles coupled to light-emitting optical waveguides,” Opt. Express 21(4), 4250–4262 (2013). [CrossRef] [PubMed]
19. P. B. Johnson and R. W. Christy, “Optical constants of the noble metals,” Phys. Rev. B 6(12), 4370–4379 (1972). [CrossRef]
