Ultrathin and Ultrasmooth Gold Films on Monolayer MoS₂

Abstract

films are also the key element of plasmonic waveguides [10,11] and hyperbolic metamaterials. [12,13] Growth of continuous and ultrathin gold films on different substrates, such as glass, silicon oxide, silicon nitride, graphene etc. is notoriously difficult due to the poor wetting of gold to these substrates. [14,15] The growth kinetics of metal films is generally determined by the adsorption and diffusion behavior of metal adatoms on the substrate. A small ratio of the adsorption energy of metal adatoms on the substrate to the bulk cohesive energy of the metal and low diffusion barrier for an adatom favor the 3D island growth behavior also known as the Volmer-Weber growth mode. [16] Within the framework of this growth model, the formation of a metal film is associated with the following stages: nucleation of islands, island growth, island impingement and coalescence, percolation, and channel filling to finally form a continuous thin film. To reduce the percolation threshold of ultrathin gold films, adhesion or seed layers of Ti, Cr, Ni, Pt, or Ge are commonly used. However, these adhesion layers significantly affect the optical and electrical properties of ultrathin metal nanostructures. [17][18][19][20][21][22] Recently, the organosilane-based adhesion layers (mercaptosilanes and aminosilanes) were used for the deposition of sub-10 nm thick continuous Au films on silicon and glass surfaces. [23][24][25][26][27] However, organosilanes are not compatible with nonoxidized silicon surfaces and poorly compatible with standard lift-off procedures, that imposes severe limitations to their use as adhesion layers. [21] Adhesion layers based on organosilanes are also inefficient for the deposition of atomically thin metal films [14] and does not move us closer to the deposition of 2D layers from bulk plasmonic metals. Actually, the latter seems now as impossible as the deposition of atomically thin carbon films had been considered before 2004. [28] In the present paper, we propose the use of MoS 2 monolayer as an entirely new type of "universal" (i.e., it can be transferred to any arbitrary substrate) [29][30][31] adhesion layer for ultrathin (<10 nm) high-quality continuous gold films. To test the feasibility of this idea, we deposited ultrathin gold films of different thicknesses onto monolayer MoS 2 , grown on silicon dioxide substrates (Figure 1a), and studied their structural and optical properties.An electron beam evaporator Nano Master NEE-4000 was used to deposit Au films on top of atmospheric pressure CVD (APCVD)-grown full area coverage MoS 2 monolayers on silicon wafers with a 285 nm thick SiO 2 coating (from 2D semiconductors Inc.). The deposition was performed at Sub-10 nm continuous metal films are promising candidates for flexible and transparent nanophotonics and optoelectronic applications. In this article, it is demonstrated that monolayer MoS 2 is a perspective adhesion layer for the deposition of continuous conductive gold films with a thickness of only 3-4 nm. Optical properties of continuous ultrath...

Detailed Experiment-Theory Comparison of Mid-Infrared Metasurface Perfect Absorbers

“…In Fig. 4, we listed several conventional dielectric parameters from popular data sets by Palik, Mcpeak, Olmon, Werner, Yakbousky, Lermarchand, Jhonson and Christy [7,8,[10][11][12][13][14]. These parameters show variation in wavelength and spectral shape.…”

“…Drude parameters of the permittivity of Au were obtained form the below listed references. The parameters of D = 600 nm , P = 1200 nm, t i = 50 nm were used: (a) Palik [7], (b) McPeak [10], (c) Olmon [11], (d) Werner [12], (e) Yakubovsky [13], (f) Lemarchand [14], (g) Babar [15] and (h) Jhonson and Christy [8]. The FDTD simulation results by (a) Palik parameters from the FDTD Lumerical software database (a default) were used for the Drude model calculations (see Table).…”

Detailed Experiment-Theory Comparison of Mid-Infrared Metasurface Perfect Absorbers

“…It is well known that the susceptibility tends to show different values according to the sample fabrication methods, thickness of the film and analytical methods of characterisation. In Figure 4, we listed several conventional dielectric parameters from popular data sets by Palik, Mcpeak, Olmon, Werner, Yakbousky, Lermarchand, Johnson and Christy [15,16,[18][19][20][21][22]. These parameters show variation in wavelength and spectral shape.…”

“…The results are shown in Figure 4. The parameters of D = 600 nm , P = 1200 nm, t i = 50 nm were used: (a) Palik [15], (b) McPeak [18], (c) Olmon [19], (d) Werner [20], (e) Yakubovsky [21], (f) Lemarchand [22], (g) Babar [23] and (h) Johnson and Christy [16]. The FDTD simulation results by (a) Palik parameters from the FDTD Lumerical software database (a default) were used for the Drude model calculations (see table).…”

Surface-Enhanced Raman Spectroscopy on Hybrid Graphene/Gold Substrates near the Percolation Threshold

“…These films are usually prepared by the high-vacuum evaporation of a noble metal (gold, silver, or copper) onto a supporting substrate (semiconductor or dielectric). The presence of two-dimensional materials on the surface of the substrate leads to a change in the kinetics of metal growth, which affects the optical properties of the thin metal films [26,27] and the critical thickness that determines the percolation threshold. These changes provide additional options for modifying SERS substrates.…”