Polarized surface-enhanced Raman spectroscopy of suspended carbon nanotubes by Pt-Re nanoantennas

Christian Bäuml, Tobias Korn, Christoph Lange, Christian Schüller, Christoph Strunk, and Nicola Paradiso

Phys. Rev. B 96, 035408 – Published 7 July 2017

Abstract

We present optical nanoantennas designed for applications that require processing temperatures larger than $800^{\circ} C$ . The antennas consist of arrays of Re/Pt bilayer strips fabricated with a lift-off-free technique on top of etched trenches. Reflectance measurements show a clear plasmonic resonance at approximately 670 nm for light polarized orthogonal to the strip axis. The functionality of the antennas is demonstrated by growing single-walled carbon nanotubes (CNTs) on top of the antenna arrays and measuring the corresponding Raman signal enhancement of selected CNTs. The results of the measurements are quantitatively discussed in light of numerical simulations which highlight the impact of the substrate.

Received 17 March 2017

DOI:https://doi.org/10.1103/PhysRevB.96.035408

Physics Subject Headings (PhySH)

Growth Nanoantennas Plasmonics Surface plasmons

Nanotubes

Surface-enhanced Raman spectroscopy

Condensed Matter, Materials & Applied PhysicsAtomic, Molecular & Optical

Authors & Affiliations

Christian Bäuml, Tobias Korn, Christoph Lange, Christian Schüller, Christoph Strunk, and Nicola Paradiso^*

Institut für Experimentelle und Angewandte Physik, University of Regensburg, 93053 Regensburg, Germany

^*nicola.paradiso@physik.uni-regensburg.de

Article Text (Subscription Required)

Click to Expand

Supplemental Material (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand

Issue

Vol. 96, Iss. 3 — 15 July 2017

Reuse & Permissions

Access Options

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
(a) Sketch of the sample in top view. (b) Side-view scheme. Carbon nanotubes (CNTs) start growing from clusters of catalyst particles (brown circles) deposited in the middle of the rectangular structure. The antenna structures are defined by evaporating a Re-Pt bilayer on top of an array of etched trenches. (c) Scanning electron microscope (SEM) micrograph of a structure after the chemical vapor deposition (CVD) step at $850^{\circ} C$ . (d) Zoom in on the right antenna array. The effect of the high temperature of the CVD step is limited to a modest ( $\approx 5$ nm) corrugation of the metal surface.
Reuse & Permissions
Figure 2
Reflectance maps measured on Re/Pt antennas for incident wavelength $λ_{L} = 633$ nm and for different polarization directions, indicated by the red arrow. The maps have been acquired by scanning the laser spot over the sample with steps of 200 nm. The reflected signal is maximal for polarization direction parallel to the array axis and minimal for polarization perpendicular to the array axis. This effect is not observed for $λ_{L} = 532$ nm, where no plasmonic resonance is excited [29]. The reflectance is normalized to that of the substrate.
Reuse & Permissions
Figure 3
Raman spectra ( $λ_{L} = 633$ nm) for CNTs overgrown on optical antenna arrays (left panels) and corresponding AFM topography images (right panels). Red and blue curves refer to measurements on CNT portions overgrown on antennas or lying on the substrate, respectively. The red and blue circles indicate the position and the size of the focused light spot for the corresponding measurement. The white arrows indicate the polarization directions. In (a) and (b) the antenna strips consist of a Re/Pt bilayer. In (c) antenna strips consist of Ti/Pt. Compared to the structures above, the corrugations due to the CVD process are much more pronounced. All the Raman spectra have been obtained after 30 s of integration, except the bare CNT signal in (a), where the time was 10 times longer and the signal was then divided by a factor 10.
Reuse & Permissions
Figure 4
(a) Sketch of the geometry considered in the analytical calculations plotted in (b). An infinitely long Pt strip of elliptical section is illuminated from above by a plane wave of amplitude $| E_{0} |$ . The incident field polarization is perpendicular to the strip axis. The vertical axis $t = 25$ nm equals the thickness of the strips in our experiment. The point $Q$ indicates the position where the field strength is calculated. (b) Corresponding color plot of the electric field intensity calculated in the point $Q$ at 5 nm from the strip edge. The field is plotted as a function of strip width $w$ and excitation frequency $ν$ . (c) In a complementary calculation we consider the effect of the substrate alone (see text). The substrate consists of a doped Si wafer capped with a 470-nm-thick ${SiO}_{2}$ layer. (d) The corresponding color plot for the electric field amplitude at the top surface of the ${SiO}_{2}$ layer. (e) Sketch of the geometry used in our finite-difference frequency-domain (FDFD) simulation. We used the following parameters: gap $g = 40$ nm, etching depth $e = 100$ nm, ${SiO}_{2}$ layer thickness $s = 470$ nm, strip thickness $t = 25$ nm (the Re and Pt layers are 7 and 18 nm thick, respectively) The simulation box width and height are $b = 240$ nm and $h = 1000$ nm, respectively. The corners of the metal strip have been rounded (curvature radius $r = 5$ nm). The width $w$ is swept from 10 to 500 nm and the field frequency $ν$ is swept from 200 to 1000 THz. (f) Electric field amplitude calculated by FDFD in the point $Q$ located along the strip axis at 5 nm from the metal edge. (g) Graph of the electric field amplitude as a function of the incident frequency for $w = 100$ and 200 nm, corresponding to the two horizontal cuts in (f).
Reuse & Permissions

Physical Review B

covering condensed matter and materials physics