Reproducible Ultrahigh Electromagnetic SERS Enhancement in Nanosphere-Plane Junctions

Surface enhanced Raman scattering (SERS) in nanoscale hotspots has been placed great hopes upon for identification of minimum chemical traces and in-situ investigation of single molecule structures and dynamics. However, previous work consists of either irreproducible enhancement factors (EF) from random aggregates, or moderate EFs despite better reproducibility. Consequently, systematic study of SERS at the single and few molecules level is still very limited, and the promised applications are far from being realized. Here we report EFs as high as the most intense hotspots in previous work yet achieved in a reproducible and well controlled manner, that is, electromagnetic EFs (EMEF) of 10^9~10 with an error down to 10^+/-0.08 from gold nanospheres on atomically flat gold planes under radially polarized (RP) laser excitation. In addition, our experiment reveals the EF's unexpected nonlinearity under as low as hundreds of nanowatts of laser power.


2
In the last decade, SERS detections of single small molecules in aggregates of metallic nanoparticles have been confirmed by the bi-analyte method, despite the extreme randomness of hotspot intensities and EFs 8,9 . It has been shown that the EFs vary from around 10 4 to over 10 10 , the 0.0003% most intense hotspots contributing 7% of the overall SERS signal 10 . It has also been pointed out that the most intense hotspots are required for detection of single small molecules with non-resonant Raman scattering cross sections as small as 10 -29 to 10 -30 cm 2 sr - 1 11 . To resolve the extreme randomness of EFs so as to achieve efficient and systematic study of molecular dynamics, well-controlled fabrication of SERS substrates has been studied extensively [12][13][14][15][16] . For example, electron beam writing of sub-5 nm gap optical antennas has been demonstrated recently, which nevertheless is no longer reproducible at such a small length scale 12,15,16 . An alternative approach that is directly related to our work in this paper is the nanoparticle-plane junction 2,5,17,18 , with reported experimental SERS EFs limited to about 10 6~8 , and with its reproducibility shown only after averaging tens of hotspots. Meanwhile, there has been great progress in tip-enhanced Raman scattering (TERS) in recent years 6,[19][20][21][22] . But due to its inherent limited EFs, TERS has only been used to detect single molecules with large Raman scattering cross sections.
In this paper, we report another kind of SERS experiment to achieve both reproducible and ultrahigh SERS EFs for the first time, as shown in Fig. 1a. A chemically synthesized 60 nm gold nanosphere is on top of a 200 nm thick atomically flat gold plane, and a RP He-Ne laser beam at 633 nm is focused by an objective with a numerical aperture (NA) of 0.9 to excite the nanosphere 23 . A monolayer of malachite green isothiocyanate (MGITC) molecules is coated on the surface of the nanosphere whose Raman scattering is collected by the same focusing objective. The nanosphere pairs with its mirror image to form a vertically oriented and vertically polarized optical antenna. As shown in Fig. 1b, the localized surface plasmon resonance (LSPR) 3 spectrum of one of the antennas is measured by collecting its side scattering of a supercontinuum source focused through the objective. The laser wavelength and three of the strongest Raman peaks of MGITC at 1180, 1370 and 1618 cm -1 are labeled to show that they all fall within the LSPR resonance. In Fig. 1c Fig. 1e. The EF is defined by comparing with an imaginary experiment in which the molecule is measured in air using a linearly polarized (LP) laser beam and the same focusing objective. Details of EF calculation are described in Supplementary Methods. In the calculation, the hotspot area, Ahotspot, is taken to be 9.3 nm 2 according to the FDTD result, which will be discussed later. We attribute the calculated EF to electromagnetic effects, since in both the antenna experiment and the bare gold plane experiment, the thiol group (-SH) of MGITC forms a covalent bond with the gold surface so that they are expected to have chemical EFs close to each other. comparable to the highest in previous reports on random aggregates 3,7,10,11 . In addition, thanks to the low laser power, the SERS signals were stable for more than five minutes without any obvious evidence of molecule degradation.
In the above, we have used MGITC as the probe molecule due to its large resonant Raman scattering cross section at 633 nm, so that the Raman signals from the bare gold plane can be measured to calculate EMEF. We also measured a monolayer of non-resonant small molecules, 4-nitrobenzenthiol (4NBT), from twenty antennas, as shown in Fig. 3. Not only can we observe clear SERS signals from the -NO2 stretching mode at 1336 cm -1 under 300 nW laser 7 power and 4 s integration time, but a considerably better reproducibility than that of MGITC which is 10 ±0.08 . We suppose the higher reproducibility to be the consequence of the molecules' better chemical stability when they are non-resonant with the laser 25 and having a larger number of smaller molecules in each hotspot. The value of Ahotspot needs further discussion. Values from less than 1 nm 2 to several tens of nm 2 have been used in the literature. Ultra-small hotspots seem to be evidenced by ultrahigh resolution TERS mapping experiments, both under ultrahigh vacuum and low temperature and in ambient conditions 6,22 . The sub-nm TERS hotspots were related to the nonlinear dependence of TERS intensity on laser power, which was suggested to result from stimulated Raman scattering (SRS) 6 . In our experiment, a strong nonlinear dependence of the SERS intensity on the laser power has also been observed, as shown in Fig. 4. However, we can exclude the possibility of SRS effect by working with a low laser power, as explained in the following. 300 nW at 633 nm 8 corresponds to 9.5×10 11 photons per second, and the plasmon life time in the antenna is 5.3 fs according to the LSPR bandwidth in Fig. 1b, so that no more than 5.0×10 -3 plasmons are simultaneously confined in the antenna under 300 nW laser power. Therefore SRS by the plasmons is much weaker than the spontaneous Raman scattering 26 . The origin of nonlinearity is not clear to us at the moment, and we don't know whether it implicates smaller hotspots and higher EFs than we have estimated in this paper.   Supplementary Fig. S5 for further discussion. Third, the EFs roll over at sub-W laser powers, as shown in Fig. 4. The reproducibility significantly worsens in the roll-over regime, therefore low laser power operation is critical, which in turn requires high sensitivity.
In conclusion, by focusing an RP laser beam onto the gold nanosphere -atomically flat gold plane junctions, we have obtained ultrahigh SERS EFs that are quite uniform between different hotspots. Together with the benefits of low power operation, this method should facilitate systematic study of nanoscale molecular behavior by Raman spectroscopy 28 . It also provides a sensitive and reproducible probe for exploring the physics of nanoscale hotspots, e.g. nonlinearity 6 , nonlocality and quantum tunneling 16,29,30 . In the future, we will integrate the nanosphere with the tip of a scanning force microscope for imaging and precise gap size control.

Sample preparation:
The antennas coated with a monolayer of MGITC were prepared as follows 17  The antennas coated with a monolayer of 4NBT were prepared as follows. First, 0.5 mL of 6.5×10 9 /mL gold nanosphere ultra-purifed water solution was added to 0.5 mL of 4 μM 4NBT (Sigma-Aldrich) water solution and mixed for 2 hours at room temperature. Then the functionalized gold nanosphere solution was drop-casted onto the gold planes. Next the samples were dried under a stream of nitrogen.
Dual-beam focused ion beam (FIB) milling was used to fabricate the double rings and align the rings' centers to the nanospheres.

Optical measurement:
Raman scattering was measured as follows. A He-Ne laser working at 632.8 nm and TEM 00 mode was used to excite the molecules. The laser beam passed through a liquid crystal polarization converter (ARCoptix) and was converted to an RP state of polarization. The RP laser beam was focused onto the samples through a long working distance 100× Plan Apo objective, whose NA is 0.9. Reflection from the sample, including Raman scattering, was collected by the same objective, passed through a long-pass filter, and detected by a monochromator installed with an electron multiplying CCD (EMCCD) detector. LSPR was measured as follows. A super continuum source was focused onto the samples through the same 100× objective. The scattered light was collected outside the NA of the objective with a lens whose NA is 0.15. The collecting lens focused the scattered light into a fiber-bundle, which was directed to the monochromator and EMCCD detector. The power of the super continuum source was carefully decreased by neutral density filters in order not to damage the samples. 11 The small scattering cross section of the antennas and the large reflection off the gold plane render it extremely difficult to find the nanospheres under optical microscopes without special methods. The same 100× objective was used as part of a home-built microscope to observe the nanospheres. A spatial filter blocked the central part of the objective's entrance pupil so that the nanospheres were illuminated at an inclined angle. The nanospheres appear as dark spots on a bright background, due to the antennas' absorption and scattering of the inclined illumination. In addition, FIB milled position markers were made on the gold planes, and SEM images were taken to compare with the optical microscopy images, so that the nanospheres can be identified repeatedly.
We have selected those gold nanoparticles with spherical shapes under SEM for optical experiments. Around 10% of the gold nanoparticles have irregular non-spherical shapes. Otherwise, we have not intentionally excluded any nanospheres for SERS EF reproducibility characterization.

FDTD simulation
The FDTD simulations were done with Lumerical FDTD Solutions. The nanosphere-plane junction structure is excited by a broadband total-field scattered-field source, which is a p-polarized planewave at 30-to-normal incidence. The boundary conditions are perfectly matched layers except for one mirror symmetry plane across the center of sphere. The finest grid size of the mesh is 0.05 nm in and near the junction gap, and increases to 4 nm at away from the junction gap.

Nanosphere-Plane Junctions
Jing Long, Hui Yi, Hongquan Li and Tian Yang Figure S1 FDTD simulation of |E z 2 | at the center of the antenna's junction gap. The antenna in Fig. 1. is used.

Hotspot intensity spectrum by FDTD simulation
The illumination is an around 30-to-normal p-polarized broadband planewave. The intensity is normalized by the |E z 2 | of illumination.

Supplementary Methods: EMEF calculation
The EMEFs of our experiments are calculated by equation (1). The first line of equation (1)  of the hotspot and the laser focal spot, respectively, so that I/A is proportional to Raman scattering power per molecule. Note that Ahotspot is the area of |E 4 | since SERS EMEF in an LSPR hotspot is proportional to |E 4 |, while ARP = 0.08 m 2 is the area of |E 2 | since Raman is a linear process by itself [1,2]. Due to the mirror effect of the gold plane, the transverse E is weak and only Ez is considered in the estimation of A's. The 2 4 factor counts for the EM enhancement contributed by the bare gold plane compared to the imaginary in-air experiment, which includes the mirror effect which increases the excitation |Ez 2 | by a factor of around 2 2 , and the Purcell effect which increases the Raman emission rate by another factor of around 2 2 . |ERP 2 |/|ELP 2 | equals the ratio between the |E 2 | values at the center points of the RP and LP focal spots, which is 3 theoretically calculated to be 0.89. A line-scanning of the FDTD simulation result of the hotspot ( Fig. 1c) is shown in Fig. S2, which gives a FWHM hotspot diameter of 3.45 nm and Ahotspot of 9.3 nm 2 .

Comparison between RP and LP focal spots
RP and LP laser beams pointing in the z direction are focused through an objective with NA=0.9.
The vectorial profiles of the focal spots are experimentally characterized by raster scanning a gold nanosphere on a silica aerogel substrate and measuring the scattered far field, following Ref [3]. The theoretical calculation results are also presented, following Ref [4]. The vectorial profiles shown in Fig. S4 indicate much larger |Ez 2 | in the RP focal spot than in the LP focal spot.
5 Figure S4 Vectorial profiles of RP and LP focal spots. The first row is RP focal spot profiles. The second row is LP focal spot profiles. Each column is the profile of either |E y 2 | or |E z 2 |, as labeled. Laser wavelength is 632.8 nm.
Focusing objective NA=0.9. All images are 1.2×1.2 μm 2 . a, Experimental results. b, Theoretical results. The intensities are normalized to the maximum |E 2 | of the respective focal spots.

Effects of nanosphere-plane interfaces
The TEM image of a 60 nm gold nanosphere coated with a monolayer of MGITC is shown in Fig. S5a. It is a polyhedron. The contact between the polyhedron and the plane may be in the form of facet, edge or apex. We have observed significant deviation between the LSPR spectra of different antennas, as shown in Fig. S5b and c. Three out of twenty antennas show double peaks in their LSPR spectra, as the orange curve in Fig. S5c. This is reported to result from strong charge concentration and plasmon coupling at the junction, which is sensitive to the morphology of the interface [5,6]. On the other hand, the high reproducibility of SERS EFs implicates that the values of Ahotspot may not vary a lot between different antennas.
6 Figure S5 a, TEM image of a 60 nm gold nanosphere coated with a monolayer of MGITC. b, LSPR spectra of twenty antennas, normalized to the same height. These antennas are the same as in Fig. S3a. c, Some representative LSPR spectra from b.