Gold platelets for high-quality plasmonics

In recent years, plasmonic nanostructures have received great attention due to their impressive capacity to improve light-matter interaction at many length scales, for example in solar cells [1], solar-driven water splitting [2] or ultra-sensitive sensing [3]. Furthermore, plasmonic nanoantennas are ideal structures for the analysis of single emitters with unprecedented resolution far below the diffraction limit, such as in quantum dots or molecules [4] and [5]. Finally, plasmonic circuits are able to perform logical operations with photons only [6]. One of the most intensively studied materials for plasmonics is gold, due to its plasmon resonance in the visible regime and its stability against oxidation in ambient air, contrary to other (plasmonically more favorable) metals such as aluminum or silver.

To prepare antennas with reliable properties, a high fabrication accuracy in the range of 1 nm is required. Typically, vapor-phase deposited gold layers are used, but their intrinsic structure is far from being ideal: (i) deposited Au layers are nano-crystalline consisting of grains with mean sizes of 10–30 nm; (ii) such layers always show a certain roughness. Both, grain boundaries and roughness negatively influence the standard processes of nanopatterning, namely electron beam lithography or focused ion beam (FIB) milling, because different crystal orientations and topography show different etching rates. Thus, the quality of plasmonic nanostructures in gold is limited by the material's quality prior to patterning and not by the finesse of the nano-structuring procedure itself. As plasmonic structures often need geometrical accuracies of <10 nm over distances of many μm, evaporated gold has been found to be a bottleneck in fabrication. However, the capabilities of ion-beam assisted nanostructuring have increased dramatically in the last years. Instead of the commonly used gallium, FIBs with noble gases such as neon or helium enable a much higher resolution and smaller structure sizes of only a few nanometers. Consequently, the need for materials with higher quality has also increased.

A few years ago this problem was solved by the group of Bert Hecht [7]. They opened the way to high-quality plasmonics with outstanding spatial resolution and reproducibility. The success story began in 2004, when wet-chemical synthesis routines for large, ultraflat and supposedly single-crystalline gold platelets were published [8][9] and [10]. In the following years, chemists improved the synthesis toward mass production and larger platelet sizes, still unnoticed by the nano-optics community. In 2010, Huang et al. were the first to use the platelets as a superior substrate for nanoplasmonic structures and they demonstrated their improved optical properties in comparison with deposited layers [7]. Since then, gold platelets have become a prominent material for high-quality plasmonics [11] and [12].

Recently, our group has investigated the structural and optical properties of gold platelets in yet unknown detail [13]. We found that the platelets can contain twin boundaries parallel to the large area {111} surface which exist throughout the whole particle. However, these twins are not expected to negatively influence the typical nano-patterning or the plasmonic properties. We also derived the complex dielectric function of single platelets by using micro-ellipsometry and showed for the first time that their optical properties agree with single crystal bulk measurements. Finally, we have evaluated a FIB-based thinning procedure, proving the possibility to create monocrystalline gold layers as thin as 10 nm.

The cover image of this issue of Materials Today shows an agglomeration of gold particles and platelets with sizes of up to 30 μm and triangular or hexagonal shapes. The here shown platelets mostly have a thickness between 400 and 800 nm which is ideal for the realization of 3-dimensional plasmonic antennas. For the fabrication of in-plane plasmonic applications, platelets with thicknesses of 50–100 nm are typically used. Usually, flakes that lie flat on a substrate surface are more desirable, but from an esthetical viewpoint, such an agglomeration is interesting as well. This micrograph was acquired with a field-emitter scanning electron microscope (MIRA3 from TESCAN) equipped with four different electron detectors at 10 kV acceleration voltage and a sample tilt of 50°. We created a so-called virtual detector which is composed of three of the detectors, namely the backscatter electron detector (BSE), the In-Beam secondary electron (IB-SE), and the chamber secondary electron detector (SE). The detectors show different signals due to distinct interaction mechanisms, shadowing and collection geometry. All three signals were simultaneously acquired and combined; the BSE signal was assigned to the red channel, the IB-SE signal is represented by green and the SE signal contributed the blue channel to the final RGB color image, respectively. As a result, the out-of-microscope image was already colored.


We would like to thank Muhammad Bashouti and Ahmed Salaheldin for the platelet synthesis. Financial support by the EU-FP7 project UnivSEM (Grant Agreement n°280566), by the DFG Research Training Group GRK1896 and by the Cluster of Excellence EXC315 “Engineering of Advanced Materials” is gratefully acknowledged.

Further reading

[1] M.A. Green, S. Pillai
Nat. Photonics, 6 (2012), pp. 130–132

[2] S.C. Warren, E. Thimsen
Energy Environ. Sci., 5 (2012), pp. 5133–5146

[3] J. Langer, S.M. Novikov, L.M. Liz-Marzán
Nanotechnology, 26 (2015), p. 322001

[4] L. Rogobete, et al.
Opt. Lett., 32 (2007), p. 1623

[5] B. Hoffmann, et al.
Nanotechnology, 26 (2015), p. 404001

[6] H. Wei, et al.
Nat. Commun., 2 (2011), p. 387

[7 ]J.-S. Huang, et al.
Nat. Commun., 1 (2010), p. 150

[8] X. Sun, S. Dong, E. Wang
Angew. Chem. Int. Ed., 43 (2004), pp. 6360–6363

[9] Y. Shao, Y. Jin, S. Dong
Chem. Commun. (Camb.) (2004), pp. 1104–1105

[10] J.-U. Kim, et al.
Adv. Mater., 16 (2004), pp. 459–464

[11] P. Geisler, et al.
Phys. Rev. Lett., 111 (2013), p. 183901

[12] C.Y. Wu, et al.
Nano Lett., 11 (2011), pp. 4256–4260

[13] B. Hoffmann, et al.
Nanoscale, 8 (2016), pp. 4529–4536

Read full text on ScienceDirect

DOI: 10.1016/j.mattod.2016.03.017