Nanostructured potential of optical trapping using a plasmonic nanoblock pair Yoshito Tanaka, Shogo Kaneda and Keiji Sasaki* Research Institute for Electronic Science, Hokkaido University, Sapporo 1-2, Japan Supporting Information Experimental setup Fabrication of a gold nanoblock pair with a single nanometer-scale gap The nanoblock pair was fabricated on a glass substrate using electron beam lithography and lift-off techniques. After washing a glass substrate, the positive resist for electron beam lithography (ZEP52a, Zeon Chemicals Co.) was formed as a film via spin-coating to a thickness of 15 nm on the substrate. The spin-coated film was exposed to a given pattern using an electron beam lithography system with a high accelerating voltage (125 kv). The exposed film was then developed and rinsed using a developer solution (ZED-N5, Zeon Chemicals Co.) and a rinsing agent (ZMD-B, Zeon Chemicals Co.) specially designed for the resist. A thin film of gold was then deposited via sputtering onto the developed substrate, and the resist layer below the gold film was removed with a resist remover solution (ZDMAC, Zeon Chemicals Co.), resulting in the formation of the gold nanoblock pair on the substrate. This is known as a lift-off process. The thickness of the gold nanostructure can be controlled by adjusting the sputtering time. Because the adhesion of gold to a glass surface is weak, a thin film of titanium measuring 2 nm in thickness was first deposited as an adhesion layer to allow the preparation of mechanically strong gold nanostructure. 1
white light from Hg lamp condenser lens polarizing beam splitter cube xyz translation stage sample 164 nm Nd 3+ : YAG laser oil immersion objective ( 1, NA = 1.35) λ/2 plate beam dump DM (:dichloic mirror) DM λ/2 plate notch filter 532 nm DPSS laser CCD camera function generator (1 μs, 3 Hz) Figure S1. Schematic illustration of the experimental setup for a measurement of plasmonic trapping potential. The beam from a Nd 3+ : YAG trapping laser ( = 164 nm) was expanded by 1.5 times and re-collimated using lenses of f = 1 and 15 mm, and was directed by a oil immersion objective on the plasmonic trapping nanostructure, which sits on a xyz translation stage. In the optical path, half-wave plates were used to rotate the linear polarization. The fluorescent excitation light from a diode-pumped solid state (DPSS) laser ( = 532 nm) was directed in a similar manner without magnification; two lenses in the optical path were the same in focal length. The focal points of these two laser beams were shifted along the optical axis of the objective by using the pairs of lenses to defocus the illumination spot on the nanostructure. The green laser was modulated with TTL signals from a function generator synchronized with CCD camera. The white light from Hg lamp was used for bright-field observation of the nanostructures. Stacked notch filters filtered out the reflection and scattering of both lasers. 2
OFF Green laser modulation ON Displacement [nm] 2 1-1 -2 1 2 3 Time [s] 4 5 6 Figure S2. Position fluctuation of the trapped nanoparticle in the Supplementary Movie 1. The particle position is obtained by using a two-dimensional Gaussian fit to the fluorescent spot at each individual frame. The measured fluctuation is clearly reduced by the motion blur during a long integration time. z 15 x z y x [nm] 15 3-3 3.8-15 15 3 15 3 15 3 y [nm] 15 z x (c) -15 15 (b) -3 log E/E 2 (a) -3-15 x [nm] -2.3 15 (d) z y -3-15 y [nm] Figure S3. Calculated field intensity distributions above a model nanoblock pair with the incident polarization direction (a, b) parallel and (c, d) perpendicular to the pair axis. (a, c) x z planes and (b, d) y z planes as shown to the left side of panels. 3
.25.2 Absorption.15.1.5. 7 9 1 11 12 Wavelength [nm] Figure S4. Calculated plasmon resonance spectra of a model nanoblock pair with incident polarization direction parallel (red) and perpendicular (blue) to the pair axis. The vertical green dotted line represents the incident laser wavelength at 164 nm. The plasmon resonance with transverse polarization is about one-fifth of the resonance intensity with longitudinal polarization at incident laser wavelength of 164 nm. Thus, the absence of significant heating can be still valid even at the illumination intensity used in Fig. 4 (transverse polarization), which was 4x stronger than one used in Fig. 2 (longitudinal polarization). 4
(a) Potential Energy Thermal fluctuation < λ/2 > λ/2 k B T (b) ~ λ/2 Figure S5. Double trapping potential wells created by focusing two parallel laser beams in conventional far-field trapping. The distance between two laser beams larger (a) and smaller (b) than a half of incident laser wavelength. The two potential wells in (a) are clearly formed and, with high intensity laser power, each potential well can confine a dielectric nanoparticle with thermal fluctuations smaller than the diffraction limit. On the other hand, as shown in (b), the separation of two potential wells cannot be shortened at a distance smaller a half of incident laser due to diffraction limit because its potential shape is determined by a sum intensity of two focused beams. 5
Captions for Supplemental Movies Supplementary Movie 1. Plasmonic trapping of a 1-nm diameter polystrene sphere. At ~3 seconds, the fluorescence excitation laser is switched from CW to 3µs pulse mode. The movie shows the effect of integration time on fluctuation measurements of nanoparticle position in plasmonic trap. The moive is recorded and played at 3 fps. Supplementary Movie 2. Plasmonic trapping of a 1-nm diameter polystrene sphere with the incident polarization direction along the pair axis. The moive is recorded and played at 3 fps. Supplementary Movie 3. Plasmonic trapping of a 1-nm diameter polystrene sphere with the polarization direction perpendicular to the pair axis. The moive is recorded and played at 3 fps. 6