Recent progress in the on-surface synthesis of graphene nanoribbons (GNRs) has given access to atomically precise narrow GNRs with tunable electronic band gaps that makes them excellent candidates for room-temperature switching devices such as field-effect transistors (FET). However, in spite of their exceptional properties, significant challenges remain for GNR processing and characterization. This contribution addresses some of the most important challenges, including GNR fabrication scalability, substrate transfer, long-term stability under ambient conditions and ex situ characterization. We focus on 7- and 9-atom wide armchair graphene nanoribbons (i.e, 7-AGNR; and 9-AGNR) grown on 200 nm Au(111)/mica substrates using a high throughput system. Transfer of both, 7- and 9-AGNRs from their Au growth substrate onto various target substrates for additional characterization is accomplished utilizing a polymer-free method that avoids residual contamination. This results in a homogeneous GNR film morphology with very few tears and wrinkles, as examined by atomic force microscopy. Raman spectroscopy indicates no significant degradation of GNR quality upon substrate transfer, and reveals that GNRs have remarkable stability under ambient conditions over a 24-month period. The transferred GNRs are analyzed using multi-wavelength Raman spectroscopy, which provides detailed insight into the wavelength dependence of the width-specific vibrational modes. Finally, we characterize the optical properties of 7- and 9-AGNRs via ultra-violet-visible (UV-Vis) spectroscopy.

Since the discovery of surface-enhanced Raman scattering (SERS) 40 years ago, the origin of the “background” that is systematically observed in SERS spectra has remained questionable. To deeply analyze this phenomenon, plasmon-resonant Raman scattering was recorded under specific experimental conditions on a panel of composite multilayer samples containing noble metal (Ag and Au) nanoparticles. Stokes, anti-Stokes, and wide, including very low, frequency ranges have been explored. The effects of temperature, size (in the nm range), embedding medium (${\text{SiO}}_{\text{2}}, {\text{Si}}_{\text{3}}{\text{N}}_{\text{4}}$, or ${\text{TiO}}_{\text{2}}$) or ligands have been successively analyzed. Both lattice (Lamb modes and bulk phonons) and electron (plasmon mode and electron-hole excitations) dynamics have been investigated. This work confirms that in Ag-based nanoplasmonics composite layers, only Raman scattering by single-particle electronic excitations accounts for the background. This latter appears as an intrinsic phenomenon independently of the presence of molecules on the metallic surface. Its spectral shape is well described by revisiting a model developed in the 1990s for analyzing electron scattering in dirty metals, and used later in superconductors. The $g_s$ factor, that determines the effective mean-free path of free carriers, is evaluated, $g_s^{\mathrm{expt}}=0.33\pm 0.04$, in good agreement with a recent evaluation based on time-dependent local density approximation $g_s^{\mathrm{theor}}=0.32$. Confinement and interface roughness effects at the nanometer range thus appear crucial to understand and control SERS enhancement and more generally plasmon-enhanced processes on metallic surfaces.

Self-assemblies of gold nanocrystals (a) examined using an original spectrometry method and apparatus (b) to get a representation of their vibrational density of states (c).

Raman spectroscopy is commonly used to determine the number of layers of few‐layer graphene (FLG) samples. In this work, we focus on the criteria based on the G‐band integrated intensity and on the laser optical contrast. Limitations due to stacking order are discussed and lead to the conclusion that it is necessary to combine Raman and optical contrast to avoid misinterpretation. Both methods enable to distinguish unambiguously between single layer graphene and multilayer graphene. However, neither each method separately nor the combination of the two enable a determination of the number of layers for all possible stacking orientations. Importantly, because the two methods always significantly disagree when they fail, the comparison of the values deduced by each method allows to discriminate if the determined number of layers can be specified or not. Other important parameters (substrate, laser wavelength, objective numerical aperture) are discussed to define a reliable method to determine the number of graphene layers in FLG and its domain of validity. The proposed method that combines Raman and optical contrast measurements, carried out with a 532 nm laser and using a 100× objective with a numerical aperture of 0.9, allows the determination of the number of layers for (up to 5) FLG on the following substrates: (1) glass (soda lime glass or similar with refractive index between 1.50 and 1.55) and (2) oxidized silicon (SiO₂ on silicon, with a SiO₂ thickness of 90 ± 5 nm). The method is however limited to high quality graphene and FLG with small defect density and low residue. Copyright © 2017 John Wiley & Sons, Ltd.