Localized surface plasmons, surface plasmon

polaritons, and their coupling in 2D metallic

array for SERS
Luping Du,2 Xuejin Zhang,2 Ting Mei,1, 2* and Xiaocong Yuan3
1
Institute of Optoelectronic Materials and Technology, South China Normal University, Guangzhou 510631, China
2
Photonics Research Center, School of Electric & Electronic Engineering, Nanyang Technological University,
Nanyang Avenue, Singapore 639798, Singapore
3
Institute of Modern Optics, Key Laboratory of Optoelectronic Information Science & Technology, Ministry of
Education of China, Nankai University, Tianjin, 300071, China
*ting.mei@ieee.org

Abstract: A substrate with ease for fabrication is proposed for surface

enhanced Raman spectroscopy (SERS). A two-dimensional dielectric
grating covered by a thin silver film enables the excitation of both localized
surface plasmons (LSPs) and surface plasmon polaritons (SPPs). The finite-
difference time-domain simulation results show that the coupling between
LSPs and SPPs is able to highly improve the Raman enhancement (2 × 109
as obtained by simulation). In addition, the near-field distribution at the top
of cubic bumps along the transverse plane presents a highly regular hotspots
pattern, which is required for an ideal SERS substrate.
©2009 Optical Society of America
OCIS codes: (050.2770) Gratings; (240.0310) Thin films; (240.6680) Surface plasmons;
(240.6695) Surface-enhanced Raman scattering.

References and links

1. S. M. Nie, and S. R. Emory, “Probing single molecules and single nanoparticles by surface-enhanced Raman
scattering,” Science 275(5303), 1102–1106 (1997).
2. R. A. Tripp, R. A. Dluhy, and Y. P. Zhao, “Novel nanostructures for SERS biosensing,” Nano Today 3(3-4), 31–
37 (2008).
3. L. J. Sherry, S. H. Chang, G. C. Schatz, R. P. Van Duyne, B. J. Wiley, and Y. N. Xia, “Localized surface
plasmon resonance spectroscopy of single silver nanocubes,” Nano Lett. 5(10), 2034–2038 (2005).
4. L. J. Sherry, R. C. Jin, C. A. Mirkin, G. C. Schatz, and R. P. Van Duyne, “Localized surface plasmon resonance
spectroscopy of single silver triangular nanoprisms,” Nano Lett. 6(9), 2060–2065 (2006).
5. K. L. Kelly, E. Coronado, L. L. Zhao, and G. C. Schatz, “The optical properties of metal nanoparticles: The
influence of size, shape, and dielectric environment,” J. Phys. Chem. B 107(3), 668–677 (2003).
6. C. E. Talley, J. B. Jackson, C. Oubre, N. K. Grady, C. W. Hollars, S. M. Lane, T. R. Huser, P. Nordlander, and
N. J. Halas, “Surface-enhanced Raman scattering from individual au nanoparticles and nanoparticle dimer
substrates,” Nano Lett. 5(8), 1569–1574 (2005).
7. Y. Fleger, Y. Mastai, M. Rosenbluh, and D. H. Dressler, “Surface enhanced Raman spectroscopy of aromatic
compounds on silver nanoclusters,” Surf. Sci. 603(5), 788–793 (2009).
8. S. Shanmukh, L. Jones, J. Driskell, Y. P. Zhao, R. Dluhy, and R. A. Tripp, “Rapid and sensitive detection of
respiratory virus molecular signatures using a silver nanorod array SERS substrate,” Nano Lett. 6(11), 2630–
2636 (2006).
9. J. N. Chen, T. Martensson, K. A. Dick, K. Deppert, H. Q. Xu, L. Samuelson, and H. X. Xu, “Surface-enhanced
Raman scattering of rhodamine 6G on nanowire arrays decorated with gold nanoparticles,” Nanotechnology
19(27), 275712 (2008).
10. H. X. Xu, J. Aizpurua, M. Kall, and P. Apell, “Electromagnetic contributions to single-molecule sensitivity in
surface-enhanced raman scattering,” Phys. Rev. E Stat. Phys. Plasmas Fluids Relat. Interdiscip. Topics 62(3 3 Pt
B), 4318–4324 (2000).
11. M. Kahl, and E. Voges, “Analysis of plasmon resonance and surface-enhanced Raman scattering on periodic
silver structures,” Phys. Rev. B 61(20), 14078–14088 (2000).
12. A. G. Brolo, E. Arctander, R. Gordon, B. Leathem, and K. L. Kavanagh, “Nanohole-enhanced Raman
scattering,” Nano Lett. 4(10), 2015–2018 (2004).
13. N. Félidj, J. Aubard, G. Levi, J. R. Krenn, A. Hohenau, G. Schider, A. Leitner, and F. R. Aussenegg, “Optimized
surface-enhanced Raman scattering on gold nanoparticle arrays,” Appl. Phys. Lett. 82(18), 3095–3097 (2003).

#116455 - $15.00 USD Received 31 Aug 2009; revised 29 Oct 2009; accepted 6 Nov 2009; published 19 Jan 2010
(C) 2010 OSA 1 February 2010 / Vol. 18, No. 3 / OPTICS EXPRESS 1959
14. A. Hohenau, J. R. Krenn, F. J. Garcia-Vidal, S. G. Rodrigo, L. Martin-Moreno, J. Beermann, and S. I.
Bozhevolnyi, “Spectroscopy and nonlinear microscopy of gold nanoparticle arrays on gold films,” Phys. Rev. B
75(8), 085104 (2007).
15. N. C. Linn, C. H. Sun, A. Arya, P. Jiang, and B. Jiang, “Surface-enhanced Raman scattering on periodic metal
nanotips with tunable sharpness,” Nanotechnology 20(22), 225303 (2009).
16. J. M. Montgomery, A. Imre, U. Welp, V. Vlasko-Vlasov, and S. K. Gray, “SERS enhancements via periodic
arrays of gold nanoparticles on silver film structures,” Opt. Express 17(10), 8669–8675 (2009),
http://www.opticsinfobase.org/oe/abstract.cfm?URI=OPEX-17-10-8669.
17. D. Lynch, and W. R. Hunter, “Handbook of Optical Constants,” E. D. Palik, ed (Academic Press, New York,
1985).
18. K. A. WilElets, and R. P. Van Duyne, “Localized surface plasmon resonance spectroscopy and sensing,” Annu.
Rev. Phys. Chem. 58(1), 267–297 (2007).
19. L. L. Zhao, K. L. Kelly, and G. C. Schatz, “The extinction spectra of silver nanoparticle arrays: Influence of
array structure on plasmon resonance wavelength and width,” J. Phys. Chem. B 107(30), 7343–7350 (2003).
20. C. L. Haynes, A. D. McFarland, L. L. Zhao, R. P. Van Duyne, G. C. Schatz, L. Gunnarsson, J. Prikulis, B.
Kasemo, and M. Kall, “Nanoparticle optics: The importance of radiative dipole coupling in two-dimensional
nanoparticle arrays,” J. Phys. Chem. B 107(30), 7337–7342 (2003).

1. Introduction
Surface enhanced Raman spectroscopy (SERS) is a powerful analytical tool for chemical and
biological sensing application which provides detailed material information at molecular
level, e.g. at single-molecule sensitivity [1], for which a Raman enhancement factor at an
order of 1014 is required. As a basic requirement for sensitive SERS, an ideal substrate must
guarantee a high enhancement effect for SERS and a reproducible uniform response, and thus
needs a large area with regular hotspots, whereas easiness for fabrication is desired [2]. Since
the discovery of SERS in 1970s, researchers have made great effort theoretically and
experimentally to develop robust substrates for SERS, including single nanoparticles [1, 3–5],
nanoparticle dimers [6], clusters [7], nanorods [8] and nanowire arrays [9]. For the dimer
configuration with extremely small gap, which provides a high Raman enhancement factor to
perform the single molecule SERS, the reported electromagnetic enhancement is at an order
of 1011 [10]. This electromagnetic enhancement effect mainly comes from the excitation of
localized surface plasmons (LSPs) with strong interaction among them. Metallic periodic
structures are another type of configuration which plays a significant role as SERS substrate
in biosensing. Many works, either theoretically or experimentally, have been conducted in
recent years [11–15]. The Raman enhancement reported is mainly among 105 – 108, and is
only considering the effect of LSPs, or surface plasmon polaritons (SPPs) separately. A recent
theoretical work shows that structures comprising metal nanoparticles within periodic arrays
can produce highly regular hotspots owing to the excitation of LSPs and SPPs [16]. The
coupling between them leads to a high electric field enhancement.
In this article, we propose a structure consisting of a two-dimensional dielectric grating
covered by a silver film with a thickness of tens of nanometers for a SERS substrate. Our
structure is easy for fabrication and able to provide highly regular hotspots with much high
electric field enhancement.
2. Plasmonic Response
Figure 1 shows the proposed structure for SERS in this study. A two-dimensional array of
SiO2 cuboids with fixed size of 50 × 50 × 100 nm3 is patterned on the Si substrate, and a layer
of 40-nm-thick silver film is covered on the surface. An array of cubic bumps is formed as
viewed from the top. Obviously, such a structure can be easily fabricated using the
conventional silicon process technology. To obtain the plasmonic response of such a
structure, the DiffractMod package based on rigorous coupled wave analysis (RCWA) in the
commercial software R-Soft was applied. For water (as the environment), Si, and SiO2, the
refractive indices are set to be 1.3364, 3.99, and 1.5458, respectively; while for silver, which
is a lossy material possessing wavelength-dependent dielectric constant, the refractive index is
obtained from Ref [17]. in the visible range (400 nm-800 nm).

#116455 - $15.00 USD Received 31 Aug 2009; revised 29 Oct 2009; accepted 6 Nov 2009; published 19 Jan 2010
(C) 2010 OSA 1 February 2010 / Vol. 18, No. 3 / OPTICS EXPRESS 1960
 Fig. 1. The proposed SERS structure and its cross sectional view along the dashed line.

Figure 2 shows the absorption characteristics based on various configurations. Using the
semi-infinite water-silver flat interface model (dashed line), the reflectance at the interface for
normal incidence is: R = ((n0 – nr)2 + k2) / ((n0 + nr)2 + k2). Where n0 is the refractive index of
environment, nr is the real part of the refractive index of silver and k is the imaginary part
which gives rise to the absorption in the silver. For a 40-nm-thick silver film sandwiched
between semi-infinite water and Si (dotted line), the absorption will decrease slightly and
transmission appears due to the finite thickness of the silver film. Further, if we produce the
two-dimensional infinite dielectric grating and cover the whole surface conformally by a 40-
nm-thick silver film, multiple absorption peaks will present in the absorption spectra due to
the excitation of LSPs and SPPs (solid line). The broad ones denoted as peaks 2 and 3 in Fig.
2 correspond to the first-order LSPs and high order LSPs, respectively, whereas the sharp one
denoted as peak 1 corresponds to the SPPs.

Fig. 2. Absorption spectra based on: semi-infinite water-silver flat interface configuration
(dashed line), 40-nm-thick silver film sandwiched between water and Si configuration (dotted
line), and dielectric grating covered by 40-nm-thick silver film configuration (solid line).

At normal incidence, the SPPs can be excited in a periodic structure satisfying

2π / Λ = k0 (ε m ⋅ ε e ) / (ε m + ε e ) , (1)

where Λ is the structural period, k0 is the wavenumber of light in vacuum, and εm and εe are
the permittivity of metal and environment, respectively. The term on the right hand side of
Eq. (1) represents the wavenumber of the SPPs propagating along the single interface.
Equation (1) predicts an SPP excitation wavelength of 670 nm for Λ = 475 nm, which is
slightly longer than the numerical simulation result (637 nm). The difference is due to the

#116455 - $15.00 USD Received 31 Aug 2009; revised 29 Oct 2009; accepted 6 Nov 2009; published 19 Jan 2010
(C) 2010 OSA 1 February 2010 / Vol. 18, No. 3 / OPTICS EXPRESS 1961
modification of Eq [1]. for periodic bump structures. Figure 3 gives the near-field
distributions at different resonance conditions in the solid-line curve in Fig. 2, which is
obtained by applying the FullWave Package based on the finite-difference time-domain
(FDTD) method in R-Soft. In the simulation, the grid size is set to be 2 or 1 nm and the
periodic boundary conditions are set for X and Y directions and the perfectly matched layer
for Z direction.
If we define Eout(λ) as the local electric field at a wavelength λ, the enhancement factor for
SERS is: EFSERS(λs) = |Eout(λ)2 ||Eout(λs)2| / |E0|4, where λs is the Stokes-shifted wavelength [18].
The wavelength difference, ∆λ = λs - λ, between the incoming and scattered photons in general
is much smaller than the linewidth of a surface plasmon mode, and thus Eout(λ) is
approximately equal to Eout(λs), resulting in the commonly used expression for the
enhancement of the stocks beam: EFSERS(λs) = |Eout(λ)4| / |E0|4. Since EFSERS(λs) = EFSERS(λs, x,
y, z) is spatially varying over the particle surface, we focus on the largest value of EFSERS in
this article, like what was done in many previous literatures,. For the case of LSPs and SPPs
in Fig. 3, the Raman enhancements are 3 × 107, 4 × 108 and 3 × 108 for the high order LSPs,
the first-order LSPs and the SPPs, respectively. Apart from the high Raman enhancement, the
highly regular hotspots are obviously produced. In the following section, we take the
structural period, the film thickness and the environment refractive index (RI) as variables and
investigate their effects on the Raman enhancement.

Fig. 3. |E| distributions at the top of cubic bumps along XY plane at various resonance
conditions, i.e., the incident wavelength of (a) 450 nm corresponding to the high order LSPs,
(b) 637 nm corresponding to the SPPs, and (c) 670 nm corresponding to the first-order LSPs.
The colorbar scale is set as ln(|E|).

3. Influence of period, bump size and environment RI on Raman enhancement

Figure 4(a) shows a set of absorption spectra for the period ranging from 402 nm to 546 nm
with a step of 24 nm. The peak absorption wavelengths are denoted as λmax and their plots
against the structural period for both the LSPs and SPPs are shown in Fig. 4(b). The red-shift
trend of the resonance wavelength for SPPs for increasing period can be predicted from the
excitation condition [Eq. (1)]. As the structural period increases, the right hand side value in
Eq. (1) should be lower accordingly to satisfy the condition. Because of the dispersion
property of silver refractive index in visible range, a red shift of excitation wavelength is
required from calculation. For the LSPs, the resonance wavelength is affected by the distance
between bumps. For large structural period, meaning that bumps are separated far away, the
plasmon resonance wavelength shift results from long-range interactions. From Fig. 4(b),
LSPs resonance wavelength shows slightly blue shift as bumps distance decreases, which is in
consistence with previous works [19, 20].

#116455 - $15.00 USD Received 31 Aug 2009; revised 29 Oct 2009; accepted 6 Nov 2009; published 19 Jan 2010
(C) 2010 OSA 1 February 2010 / Vol. 18, No. 3 / OPTICS EXPRESS 1962
 Fig. 4. (a) Absorption spectra for various periods from 426 nm to 546 nm with a 24 nm
increment; (b) absorption peak positions versus different periods for LSPs (circles) and SPPs
(squares), with corresponding maximum local |E| at resonance wavelength of SPPs (triangles).
The silver film thickness is 40 nm, the bump height is 100 nm and the environment is water.

The Raman enhancement shows a great dependence on the structural period because of the
interaction between the LSPs and SPPs. In Fig. 4(b), two peaks can be found from the
|Emax(λ)| plot for the resonance wavelength of SPPs at each period (green curve with
triangles). The red (with circles) and black (with squares) curves clearly show that there is an
intersection between them, implying the occurrence of LSPs and SPPs at the same
wavelength. In this case, strong coupling between the SPPs and LSPs will occur, leading to
considerable enhancement of the local electric field responsible for the right peak in the
|Emax(λ)| curve. As for the left peak at smaller wavelength, the analysis of the |E(λ)|
distribution shows that it is from the coupling between the SPPs and the second-order LSPs.

Fig. 5. Absorption peak positions at various silver film thicknesses for LSPs (circles) and SPPs
(squares), with corresponding maximum local |E| at resonance wavelengths of SPPs (triangles).
The structural period is fixed at 475 nm, the bump height is 100 nm and the environment is
water.

Next, we consider the influence of the silver film thickness to the resonance wavelength
and Raman enhancement. For a fixed structural period, the resonance wavelength of SPPs
remains constant approximately. For the LSPs, as the film thickness increases, three aspects
should be considered for contribution to the wavelength shift. Firstly, the distance between the
bumps decreases for increasing film thickness. As stated above, this causes slightly blue shifts
of the resonance wavelength. Secondly, since the nanoparticle (bump here) is not pure metal,
as the film thickness increases, the effective refractive index of nanoparticles will change
accordingly, which induces slightly blue shifts as well. Thirdly, the size and shape of the

#116455 - $15.00 USD Received 31 Aug 2009; revised 29 Oct 2009; accepted 6 Nov 2009; published 19 Jan 2010
(C) 2010 OSA 1 February 2010 / Vol. 18, No. 3 / OPTICS EXPRESS 1963
nanoparticle play an important role in determining the plasmon resonance wavelength. As the
film thickness increases, the nanoparticle size is largen and the shape becomes more oblate,
both effects contribute to red shift of plasmon resonance wavelength [5]. The tradeoff
between these three factors results in the LSPs resonance wavelength shift in a whole. Figure
5 presents the influence of the silver film thickness to the absorption peak positions. The
factor of size and shape of the bumps seems to be dominant in our case because of the red
shift of resonance wavelength with the film thickness. As suggested in Fig. 5, the resonance
wavelengths of the LSPs and SPPs meet at silver film thickness of ~25 nm, leading to a
significantly high electric field enhancement due to strong coupling between them. As the
film thickness increases, the two resonance wavelengths deviate from each other, implying
weaker interaction between the two modes and thus the decreasing Raman enhancement. The
maximum field enhancement occurs at silver film thickness of ~29 nm, as shown in Fig. 5.
The slight shift with the 25 nm may be due to difference between the RCWA and FDTD
methods.

Fig. 6. |E| distributions under LSPs and SPPs coupling conditions in Fig. 4(b) (left), and in Fig.
5 (right). The colorbar scale in both figures is set as ln(|E|). 1 nm cell size is used in this
calculation.

Figure 6 shows the electric field distributions under LSPs and SPPs coupling conditions in
Fig. 4(b) and Fig. 5. The co-existence of the LSPs and SPPs is clearly seen in both figures.
Moreover, the electric field enhancement is greatly maximized under such a circumstance.
The Raman enhancements for these two cases are 1 × 109 and 2 × 109, respectively, with
improvement of 1 orders of magnitude compared to those in Fig. 3 and 3 orders compared to
those without any modes. Such high Raman enhancement ensures this structure as a potential
substrate for high sensitive molecule detection experiment.
Finally, the influence of environment, i.e. its RI, is considered. For the SPPs, from the
right hand side of Eq. (1), the change of the environment RI requires the adjusted permittivity
of silver to satisfy the excitation condition for a fixed structural period. According to the
dispersion of the silver’s permittivity, there will be a considerable red shift of the resonance
wavelength for increasing environment RI (black curve with squares). For the LSPs, using the
Drude model of the metal, we can obtain the following equation [5]
2
λLSP = λb (2ne 2 + 1)1/ 2 ≈ 3λb [1 + (ne − 1)], (2)
3
for the first-order LSPs, where λb is the bulk plasmon wavelength of silver and ne the
environment RI. Figure 7 presents linear relations of SPPs and LSPs absorption peak
wavelength with the environment RI. This shows that although Eq. (2) is derived from the
spherical model, a similar proportionality exists for nonspherical shapes as well. This linear
relationship between the resonance wavelengths and the environment RI allows such a

#116455 - $15.00 USD Received 31 Aug 2009; revised 29 Oct 2009; accepted 6 Nov 2009; published 19 Jan 2010
(C) 2010 OSA 1 February 2010 / Vol. 18, No. 3 / OPTICS EXPRESS 1964
structure to be applicable for biosensing application. In addition, the relationship between the
local |E(λ)|max values at the resonance wavelengths of the SPPs and the environment RI (green
curve with triangles) shows that increasing the environment RI could lead to higher Raman
enhancement.

Fig. 7. Absorption peak positions at various environment refractive indices for LSPs (circles),
SPPs (squares), and corresponding maximum local |E| at resonance wavelengths of SPPs
(triangles). The structural period is fixed at 475 nm, the film thickness is 40 nm and the bump
height is 100 nm.

4. Conclusion
We have investigated the plasmonic properties of a potential SERS substrate, on which both
the LSPs and SPPs can be excited. The FDTD simulation shows that the resonance
wavelengths of LSPs and SPPs can be effectively tuned via changing the structural period and
the silver film thickness. LSPs and SPPs can co-exist at specific wavelengths such that the
coupling between them can contribute to a considerable boost of Raman enhancement. A
Raman enhancement of ~2 × 109 is achieved in the proposed structure. In addition, the good
proportionality between the resonance wavelength of LSPs and the environment RI could lead
to a potential application for biosensing. This work illustrates a structure with large area,
regular hotspots, as well as extremely high Raman enhancement. It is a good candidate for
SERS substrates as it can be easily fabricated using conventional silicon process technology.
Acknowledgment
This work was financially supported by the Singapore National Research Foundation (CRP
Award No. NRF-G-CRP 2007-01). X. J. Zhang acknowledges the support from the National
Natural Science Foundation of China under Grant No. 10804003. X.C. Yuan acknowledges
the support from the Ministry of Science and Technology of China under Grant No.
2009DFA52300 for China-Singapore collaborations.

#116455 - $15.00 USD Received 31 Aug 2009; revised 29 Oct 2009; accepted 6 Nov 2009; published 19 Jan 2010
(C) 2010 OSA 1 February 2010 / Vol. 18, No. 3 / OPTICS EXPRESS 1965