A method for the fabrication of a nanocomposite film, e.g. a sers or lspr subtrate, and a substrate, e.g. a sers or lspr subtrate

Abstract

A substrate, e.g., a SERS or LSPR substrate, comprising an oxide film on the surface of the substrate and metallic nanoparticles deposited on the oxide film, and a method for the fabrication of such a substrate are proposed. The method comprises depositing an oxide film on a substrate, e.g., by physical vapour deposition, and depositing metallic nanoparticles on the oxide film. Prior to the deposition of the metallic nanoparticles, the oxide film is subjected to a plasma surface treatment in a reducing atmosphere (e.g., including H2). Another, independent aspect of the invention relates to substrates that can be obtained by that method.

Description

A METHOD FOR THE FABRICATION OF A NANOCOMPOSITE FILM, E.G. A SERS OR LSPR SUBTRATE, AND A SUBSTRATE, E.G. A SERS OR LSPR SUBTRATEBackground of the Invention

[0001] The invention generally relates to a substrate with a nanocomposite film and to a fabrication method of such nanocomposite films.

[0002] Raman spectroscopy is a non-destructive technique for chemical analysis within the subfield of vibrational spectroscopy (which further includes infrared and near-infrared spectroscopy). Raman spectroscopy enables a sensitive structural identification of small concentrations of chemicals. It is named after Indian physicist Chandrasekhara Venkata Raman and relies on the different functional groups, which have distinctive vibrational energies, also known as molecular fingerprints. When a molecule is hit with a laser beam, most of the light will be scattered with the same wavelength as the incident beam (this is the elastic scattering or Rayleigh Scattering) and provides no information about the molecule. However, a small fraction of the light (typically 10’7%) is Raman-scattered (inelastically scattered) so as to produce photons at different wavelengths, which depend on the chemical structure of the analyte. The inelastically scattered light can either have a higher or lower frequency than the incident beam, which is known as anti-Stokes and Stokes Raman scattering, respectively.

[0003] Surface Enhanced Raman Spectroscopy (SERS) enhances the Raman scattering of molecules supported by nanostructured materials. Nanostructured materials such as nanoparticles (NPs), roughened films or nano-patterned substrates, enhance the Raman signals of analytes through surface plasmon enhancement or by chemical contributions [1] (the list of references in square brackets is provided at the end of the description).

[0004] Commercial SERS substrates are complex to produce and / or show limited mechanical and chemical stability. A new method that allows the development of SERS substrates comprising metallic nanoparticles deposited on oxide coatings is presented herein. Substrates obtained by the method have been characterized and are also presented herein.

[0005] The paper by Diogo Costa et al. , “Development of biocompatible plasmonic thin films composed of noble metal nanoparticles embedded in a dielectric matrix toenhance Raman signals” Applied Surface Science, Volume 496, 2019, 143701 , ISSN 0169-4332, https: / / doi.Org / 10.1016 / j.apsusc.2019.143701 , discloses the production of nanocomposite thin films, composed of noble nanoparticles embedded in a dielectric matrix, to serve as plasmonic platforms for detection of molecules using SERS. Au- AI2O3, Au-TiCh and Ag-TiO2 thin films systems were deposited by reactive DC magnetron sputtering, followed by a thermal treatment at different temperatures to promote the growth of the nanoparticles. Plasma treatments were carried out after annealing of the Au-TiCh and Ag-TiCh nanostructures, i.e. , after the deposition steps. The effect of these plasma treatments was that some of the embedded nanoparticles became more exposed to the environment, which enhanced Raman signals of R6G molecules.

[0006] The paper by M.S. Rodrigues et al., “Optimization of nanocomposite Au / TiO2 thin films towards LSPR optical-sensing”, Applied Surface Science, Volume 438, 2018, pages 74-83, ISSN 0169-4332, https: / / doi.Org / 10.1016 / j.apsusc.2017.09.162, reveals that post-deposition plasma treatment with Ar leads to a gradual blue-shift of the LSPR absorption band. According to the paper, this demonstrates the sensitivity of the films to changes in the dielectric environment surrounding the NPs and the exposure of the Au nanoparticles.

[0007] The paper by Sheau Wei Ong et al., “Optical and chemical stability of sputtered-Au nanoparticles and film in ambient environment”, Applied Surface Science, Volume 488, 2019, Pages 753-762, ISSN 0169-4332, https: / / doi.Org / 10.1016 / j.apsusc.2019.05.233, discloses that gold sputtered on quartz can exhibit particle- or film-like morphologies. For Au coverage less than equivalent thickness of 3 nm, well-dispersed Au nanoparticles (Au NPs) form. The size of the Au NPs influences the colour, as does oxidation. It was shown that when the amount of sputtered Au increases, a labyrinth-like network of coalesced Au-particles forms before film-like morphology eventually emerges.Summary of the Invention

[0008] In a first aspect, the invention relates to a method for the fabrication of a nanocomposite film, e.g., a SERS or LSPR substrate. The method comprises depositing an oxide film on a substrate by physical vapour deposition (PVD) and depositing metallic nanoparticles on the oxide film. Prior to the deposition of themetallic nanoparticles, the oxide film is subjected to a plasma surface treatment in a reducing atmosphere, e.g., an atmosphere including H2 (up to 100% H2).

[0009] It has been found that the plasma surface treatment in a reducing atmosphere significantly modifies the surface structure of the nanocomposite film and thus its properties. For the sake of brevity, the expression “treated surface” or “treated sample” will be used herein for an oxide surface or sample, respectively, that has undergone plasma surface treatment in a reducing atmosphere. In contrast, an oxide surface or sample that has not undergone plasma surface treatment in a reducing atmosphere will be referred to as an “untreated surface” or “untreated sample”, respectively. XPS (X-ray photoelectron spectroscopy) measurements show that metal deposited on a treated surface has a tendency to more “vertical” growth (perpendicular to the substrate), whereas metal deposition in comparable conditions on an untreated surface results in lower height but higher coverage, i.e., a more “horizontal” growth (parallel to the surface). In other words, the nanoparticles that form on a treated surface are taller than those that form on an untreated surface. AFM (atomic force microscopy) measurements confirm that on a treated surface, a rougher, pointier takes place in the early stage of metal nanoparticle deposition. UV-Vis (ultraviolet-visible) spectroscopy measurements indicate that significantly more metal can be deposited on a treated surface than on an untreated surface before the metal nanoparticles form a percolated network, i.e., before a continuous metal film is obtained.

[0010] The oxide film may be a thin film, i.e., a layer of material having a thickness ranging from the height of a monolayer to several micrometres, e.g., 100 pm. It should be noted, however, that the oxide film could also be thicker. The oxide film may preferably comprise an amorphous oxide, but crystalline oxides are not excluded.

[0011] The oxide film may be deposited by any suitable process, e.g., CVD (chemical vapour deposition, PECVD (plasma-enhanced CVD), thermal oxidation, sol-gel deposition, etc. Preferably, the oxide film may be deposited by a PVD (physical vapour deposition) process, e.g., (magnetron) sputtering, pulsed laser deposition, etc. Particularly preferred deposition techniques include reactive sputtering, high-power impulse magnetron sputtering (HiPIMS), and reactive HiPIMS (R-HiPIMS).

[0012] The oxide film may be deposited by reactive sputtering, e.g., R-HiPIMS, in an atmosphere including O2.

[0013] The oxide film may comprise or consist of one or more of AI2O3, TiCh, CuO, WO3, and SiC>2, preferably one or more of AI2O3, CuO, WO3, and SiO2. Most preferably, the oxide film may comprise or consist of AI2O3. Other oxide films (e.g., HfO2, ZrO2, ZnO2, etc.) are not excluded. The oxide film is preferably an electrical isolator. Metal oxide films may be preferred for certain applications.

[0014] The metallic nanoparticles preferably comprise or consist of plasmonic metal nanoparticles, so as to form a plasmonic metal nanostructure on the oxide film. The metallic nanoparticles may, e.g., comprise or consist of one or more of Au, Ag, Cu, Pt, Rh and Pd.

[0015] The metallic nanoparticles may be particles having dimensions in the range from 1 to 100 nm, more preferably in the range from 1 to 50 nm and even more preferably in the range from 1 to 30 nm, in at least two mutually orthogonal directions. The measurement of dimensions may be made by SEM (scanning electron microscopy) and / or AFM, using direct and / or indirect measurements.

[0016] The reducing atmosphere of the plasma surface treatment may comprise a reducing gas, such as, e.g., NH3, CH4, or H2, H2 being most preferred. The plasma surface treatment may be carried out in an atmosphere that comprises a mixture of H2 (or another reducing gas) and a gas selected, e.g., from He, Ar, Ne, Kr and Xe, Ar being most preferred.

[0017] The plasma of the plasma surface treatment may be generated with a microwave source. Alternatively, or additionally, an RF plasma source may be used, e.g., an ICP (inductively coupled plasma) source or a CCP (capacitively coupled plasma) source.

[0018] According to a preferred embodiment, the oxide film comprises AI2O3, the metallic nanoparticles comprise Au and the atmosphere of the plasma surface treatment comprises a mixture of H2 and Ar.

[0019] The deposition of the metallic nanoparticles on the oxide film may comprise the formation of metallic nanoparticles having an aspect ratio h / t; in the range from 0.65 to 1 .50. The parameters h and may be determined in accordance with the Height and Surface Coverage Determination Method and the Feature Size Determination Method disclosed hereinafter.

[0020] The deposition of the metallic nanoparticles on the oxide film may comprise the formation of metallic nanoparticles with an areal number density less than 6000 pm-2. The areal number density may be determined in accordance with the Areal Number Density Determination Method hereinafter.

[0021] The deposition of the metallic nanoparticles on the oxide film may comprise the formation of a surface coverage of the metallic nanoparticles on the oxide film lies in the range from 30% to 50%. The surface coverage may be determined in accordance with the Surface Coverage Determination Method disclosed hereinafter.

[0022] The deposition of the metallic nanoparticles on the oxide film may, preferably, comprises the formation of a non-percolated network of the metallic nanoparticles on the oxide film.

[0023] The Ar / H2 ratio (vol. / vol.) for the plasma surface treatment preferably lies in the range from 5 to 100, more preferably in the range from 10 to 50 and yet more preferably in the range from 15 to 40.

[0024] In another aspect, the invention relates to a substrate, e.g., a SERS (surface- enhanced Raman spectroscopy) or LSPR (localized surface plasmon resonance) substrate, comprising an oxide film on the surface of the substrate and metallic nanoparticles deposited on the oxide film (so as to form a plasmonic metal nanostructure). The substrate may be manufactured using the method according to the first aspect of the invention. The metallic nanoparticles deposited on the oxide film may have an aspect ratio, h / ^, in the range from 0.65 to 1 .50. The parameters h and may be determined in accordance with the Height and Surface Coverage Determination Method and the Feature Size Determination Method disclosed hereinafter.

[0025] The metallic nanoparticles may have an areal number density (number of particles per unit area) less than 6000 pm-2. The areal number density may be determined in accordance with the Areal Number Density Determination Method hereinafter.

[0026] The surface coverage of the metallic nanoparticles on the oxide film may lie in the range from 30% to 50%. The surface coverage may be determined in accordance with the Surface Coverage Determination Method disclosed hereinafter.

[0027] The metallic nanoparticles preferably form a non-percolated network on the oxide film. Below the percolation threshold, the plasmonic metal nanostructure formed by the ensemble of the metallic nanoparticles has the properties of an ensemble of individual nanoparticles or small clusters thereof. Above the percolation threshold, the ensemble of the metallic nanoparticles starts to show properties of a continuous metal film. For instance, it can be observed by spectroscopy that the height of the LSPR absorbance peak first increases with deposition time. When metal deposition goes on, the absorbance at wavelengths above the peak wavelength suddenly increases and the peak becomes very broad. This phenomenon may be associated with the transition from an ensemble of individual nanoparticles or small clusters towards a continuous film (i.e. , percolation). The same transition may be observed when depositing metallic nanoparticles on an untreated surface but in this case, percolation occurs significantly earlier, i.e., at a significantly lower amount of deposited metal.

[0028] The oxide film may comprise or consist of one or more of AI2O3, TiCh, CuO, WOs,and SiCh. The oxide film may, more preferably, comprise or consist of AI2O3. The oxide film may, comprise or consist of one or more of CuO, WOs,and SiO2. Other oxide films (e.g., HfO2, ZrO2, ZnO2, etc.) are not excluded. The oxide film is preferably an electrical isolator. Metal oxide films may be preferred for certain applications.

[0029] The metallic nanoparticles preferably comprise or consist of plasmonic metal nanoparticles, so as to form a plasmonic metal nanostructure on the oxide film. The metallic nanoparticles may, e.g., comprise or consist of one or more of Au, Ag, Cu, Pt, Rh and Pd.

[0030] The metallic nanoparticles may be particles having dimensions in the range from 1 to 100 nm, more preferably in the range from 1 to 50 nm and even more preferably in the range from 1 to 30 nm, in at least two mutually orthogonal directions.

[0031] According to a preferred embodiment, the oxide film comprises AI2O3, and the metallic nanoparticles comprise Au.

[0032] In the present document, the verb “to comprise” and the expression “to be comprised of’ are used as open transitional phrases meaning “to include” or “to consist at least of’. Unless otherwise implied by context, the use of singular word form is intended to encompass the plural, except when the cardinal number “one” is used: “one” herein means “exactly one”. Ordinal numbers (“first”, “second”, etc.) are usedherein to differentiate between different instances of a generic object; no particular order, importance or hierarchy is intended to be implied by the use of these expressions. Furthermore, when plural instances of an object are referred to by ordinal numbers, this does not necessarily mean that no other instances of that object are present (unless this follows clearly from context). When this description refers to “an embodiment”, “one embodiment”, “embodiments”, etc., this means that the features of those embodiments can be used in the combination explicitly presented but also that the features can be combined across embodiments without departing from the invention, unless it follows from context that features cannot be combined.Brief Description of the Drawings

[0033] By way of example, preferred, non-limiting embodiments of the invention will now be described in detail with reference to the accompanying drawings, in which:Fig. 1 : is a schematic illustration of the method for determining the concentration profile c(z) of a sample under investigation by measuring J(E) of the sample and of a reference sample.Fig. 2: shows the XPS spectra of (a) untreated and (b) treated surfaces with their respective backgrounds caused by inelastic scattering (dashed curves).Fig. 3: shows graphical representations of data obtained from XPS peak-shape analysis of the spectra of Fig. 2 - Fig. 3 (a) shows the amount of substance (AOS) as a function of time, where the dashed lines represent linear fittings going through the origin and Fig. 3 (b) displays island height vs. surface coverage. The dashed lines in Fig. 3 (b) represent AOS isolines.Fig. 4: illustrates how the parameters roughness (w), correlation length ( ) and Hurst parameter (a) may be interpreted. Fig. 4 (a) shows the local surface morphology for surfaces with similar w but different a. Fig. 4 (b) schematically represents a rough surface with mean height (h), correlation length ( ) and roughness (w). Adapted from

[0012] ,Fig. 5: shows log-log plots of the HHCF of (a) untreated and (b) treated alumina surfaces with Au deposition times from 0 s to 80 s. The solid lines represent the fittings to Eq. 6.Fig. 6: shows the variation of (a) roughness w, (b) correlation length(c) Hurst parameter a and (d) local slope (m ~ y with deposition time. Straight lines in (a) and (b) represent linear fittings whose slopes [3 and 1 / z are the so-called growth and dynamic exponents, respectively. Dashed lines in (c) and (d) are guides to the eye.Fig. 7: shows absorption spectra of Au nanoparticles deposited on (a) untreated and (b) treated AI2O3 surfaces for the different deposition times indicated in the figure. The squares indicate the position of the SPR maxima.Fig. 8: shows absorption spectra of Au nanoparticles deposited for 60 s on (a) amorphous AI2O3 at room temperature and (b) Y-AI2O3 at 450°C with / without prior Ar / H2 treatment.Fig. 9: shows the absorption spectra of Au nanoparticles deposited for 30 s on alumina surfaces treated in different Ar / H2 plasma conditions. Fig. 9 (a) relates to an H2 flux of 1 .5 seem, Fig. 9 (b) to an H2 flux of 1 .5 seem.Fig. 10: shows Raman spectra of (a) R6G analyte (10-5mol / L) on untreated (60 s) and treated (80 s) surfaces (b) R6G analyte (10-3mol / L) on treated surfaces obtained with different Au deposition times (30 s, 50 s and 70 s). A reference spectrum measured in an area without Au NP is plotted for comparison.Fig. 11 : shows the schematic of the movement of the substrates in front of the target during the deposition of (a) alumina and (b) gold.Detailed Description of Preferred Embodiments

[0034] According to a preferred embodiment, a method for the fabrication of a SERS or LSPR substrate carrying a metal oxide film having on its surface sputtered plasmonic-metal nanoparticles is proposed.

[0035] The oxide film may be deposited by reactive magnetron sputtering of a metal target consisting of the metal (Ti, Zn, Al, etc.) participating in the formation of the oxide using a mixture of an inert gas (e.g., a noble gas, such as Ar, Ne, etc.) and O2 (reactive gas). A preferred deposition method for the oxide film may be R-HiPIMS. To avoid target poisoning, i.e. , the formation of a compound film not only on the substrate but also on the sputter target, the reactive gas flow may be dynamically adjusted so as tooperate in conditions where the target is quasi-poisoned, but the sputtering rate remains high. The oxide coating is preferably deposited in dynamic mode, wherein the substrate carries out an oscillating motion.

[0036] The plasmonic-metal nanoparticles are preferably deposited by magnetron sputtering in a non-reactive atmosphere (comprised of, e.g., a noble gas, such as Ar, Ne, etc.) The sputtering may also be carried out in dynamic mode. Both the DC and the Pulsed-DC sputtering techniques may be used.

[0037] For the fabrication of samples according to the invention, a plasma treatment of the oxide film is carried out in a reducing atmosphere, e.g., an atmosphere comprising H2, prior to the deposition of the metallic nanoparticles. Comparative samples may be obtained by carrying out the deposition steps using the same or essentially the same deposition parameters while leaving out the plasma treatment step or carrying out the plasma treatment step in the absence of reducing gas.Examples

[0038] Alumina coatings were deposited on substrates (e.g., glass or Si) in a PVD coater machine (volume of 850 litres) by reactive magnetron sputtering of an Al target (purity: 99.999%, surface: 500 cm2) using a mixture of Ar (50 seem) and O2 (7-9 seem) (purity 99.998%) under a working pressure of 2.1 x1 O’3to 2.3x1 O’3mbar. To avoid target poisoning, a feedback control system capable of promptly adjusting the reactive gas flow in response to the plasma conditions was employed. This allows operating in a non-stable situation where the target surface is quasi-poisoned, but the sputtering rate is still high. Fig. 11 shows the geometry of the setup. All the coatings were deposited in dynamic mode, which implies an oscillating motion of the substrates in front of the target, with a rotation speed of 5 rpm and an amplitude (I3>) of 70°. The following geometrical parameters were chosen: L1 = 200 mm, L2 = 85, L3 = 125 mm and L4 = 101.6 mm (4 inch). The distance between the substrate holder and the targets was 85 mm. Alumina films were deposited by HiPIMS (MELEC SIPP 2000 generator) at room temperature (20°C-25°C) operating in constant power mode of 2000 W, 1 kHz, and 3 % of duty cycle.

[0039] Gold nanoparticles were deposited on a AI2O3 layer of 7 nm. The gold target was sputtered in an atmosphere composed by Ar gas (50 seem) during 10 s to 80 s in dynamic mode (70°). The Au sputtering was carried out in DC and Pulsed-DC modesFor the investigation of the Pulsed-DC, four duty cycles of 40%, 20%, 4.1 % and 1 .5 % were selected (tonof 95 , 50, 15, and 15 ps, respectively, and tOff of 145, 200, 350 and 50 ps, respectively). The average power was around 31 W and 42-54 W for DC and Pulsed DC, respectively.

[0040] In treated samples, an Ar / H2 plasma treatment was carried out on the AI2O3 films prior the deposition of Au nanoparticles. The base pressure of the chamber was 0.005 Pa, and the pressure of the Ar / H2 mixture during exposure treatment was 0.08 Pa. The plasma treatment took place for 10-30 minutes, by applying a microwave power of 150 W. The influence of different Ar / H2 (volume / volume) ratios (33.3, 16.6) and treatment times (0-30 minutes) were explored.

[0041] Hereinafter samples TS1-TS8 designate treated samples: substrate was silicon and glass, the oxide layer was an AI2O3 layer of 7 nm, which was treated with an Ar / H2 plasma before deposition of Au nanoparticles. Samples TS1 to TS8 were obtained by sputtering an Au target in an atmosphere composed of Ar gas (50 seem) for 10 s (TS1 ), 20 s (TS2), 30 s (TS3), 40 s (TS4), 50 s (TS5), 60 s (TS6), 70 s (TS7) and 80 s (TS8) in dynamic mode. Comparative samples US1 -US6 were produced in the same conditions as samples TS1 -TS6, respectively, except that no intermediate Ar / H2 plasma treatment was carried out.Characterisation of Samples using XPS

[0042] The samples according to the examples were analysed by x-ray photoelectron spectroscopy (XPS). This is a surface-sensitive analytical technique used to study the atomic composition, chemical bonding, electronic structure, and band structure of a sample. It provides information about the surface of a material by measuring the energies of photoelectrons emitted when a material is exposed to X-rays. Additionally, the binding energy (BE) of a core-electron is a distinctive feature for elements in a specific chemical environment (chemical shift). In XPS, an ejection of a core-electron from a sample takes place when an x-ray photon (with energy hv) is absorbed by an electron of the sample with a lower binding energy (BE, the energy retaining the electron to the nucleus), leading to a phenomenon called photoionization. XPS counts the electrons at a specific binding energy emitted into a given solid angle. The shape of the peaks of the spectrum gives information about the distribution of an atom in a sample. The underlying idea is that an electron ejected from an atom that is buried inthe sample will lose energy in processes of inelastic scattering before reaching the detector. As a consequence of these processes, the intensity of the peak background increases. Therefore, a proper analysis of the peak shape can give information about the distribution of analysed atoms in a surface, such as values of coverage (c, in %) and island height (h, in nm). This method was developed by Tougaard (Tougaard, 1988, 1997, 2010) [2-4], and it is implemented in the QUASES-Analyze software (version 5.0, available from QUASES- Tougaard Inc.), which was used to evaluate the characteristics of Au clusters on AI2O3 coatings.

[0043] The measured XPS spectrum J E, fl) is a function of o the intrinsic distribution of the spectrum of a single atom F(E,fT), o the number of atoms per unit volume at depth z, denoted c(z), o and the differential inelastic scattering cross section, K(T), which is the probability that an electron shall lose energy T per unit energy loss and per unit path length travelled in the solid.In these expressions, E represents the kinetic energy of the electron, and Q is the solid angle of the emission.

[0044] For a sample having a known distribution of the atom of interest as a function of depth, c(z), the emission spectrum of a single atomcan be calculated using equation Eq. 1 , below. The signal attenuation is estimated from the inelastic scattering cross section, K(T), and the inelastic electron mean free path (IMFP).where:with co ~zP = f c(z)eiMFP.cos0 dz (Eq. 3) where T=(E-E’) is the energy loss, c(z) is the number of atoms per unit volume at depth z, Q is the angle between surface normal and the detector, z / cos0 is the travelled distance, s is an integration variable, £(s) and P(s) are the Fourier transforms of energy distribution functions, which are introduced for mathematical convenience.

[0045] The following expression for the universal cross-section K(T) may be used (valid for most metals, their oxides and alloys),with C= 1643 eV2and B = 3000 eV2[2],

[0046] Eq. 1 may be used to determine an unknown distribution of the atom of interest as a function of depth in a sample. Dropping the dependence on 12, the single-atom emission spectrum F(E) may be determined by measuring J(E) on a reference sample with known c(z). IMFP may be obtained using the TPP-2 M formula (Tanuma, Powell, Penn 2 Method) [5, 6], F(E) depends only on the atom species and is the same in the reference sample and the sample under investigation. The unknown concentration profile c(z) of the sample under investigation may then be determined by measuring J(E), using the single-atom emission spectrum F(E), J(E), IMFP and K(T) as inputs in order to solve for c(z). A graphical summary of this method is shown in Fig. 1 .

[0047] The samples under investigation have an oxide film covered with metallic nanoparticles, which may be modelled as islands having an (mean) height h and together forming a certain coverage c on the oxide surface. The vertical and horizontal characteristics of the nanoparticles are described, in this model, in terms of their (mean) height and (overall) coverage. It is worth noting that the coverage affects only the intensity of the peak, whereas the height of the islands influences not only the intensity of the peak, but also the distribution of energy in the background (i.e., the shape of the peak).Height and Surface Coverage Determination Method

[0048] Samples of substrates comprising an oxide film and metallic nanoparticles on the surface of the oxide film are subjected to XPS. A reference sample is provided in the form of a foil of a known thickness consisting of the metal of the nanoparticles. For each sample, the XPS spectrumis recorded in the same conditions.

[0049] One or more spectral peaks of the metal atoms having no significant overlap with peaks relating to the oxide are selected for the analysis. For example, for Au nanoparticles on AI2O3, the Au 4f peaks overlap with the Al 2p and Al 2s peaks and the associated backgrounds (due to the inelastic scattering), making the analysisunreliable. The Au 4d peaks may in this case selected for the analysis as they show no overlap with other peaks.

[0050] The values of coverage (c, in %) and island height (h, in nm) are determined using the QUASES software, using a simple model consisting of rectangular vertical islands including a thin film of C contamination in the top surface (0.7-1.3 nm). The transformation J(E) — F(E) by optimization of the coverage and height parameters may be considered as the removal of the background caused by the inelastic scattering. A correct fit is obtained when FCE) « 0 results for a wide energy interval 50- 150 eV on the low-kinetic-energy-side of the peak [7], and when the shape of the calculated background follows the experimental background.XPS Results

[0051] Fig. 2 shows the experimental spectra of the different samples (Fig. 2 (a): US1 - US6; Fig. 2(b): TS1 -TS8) and the Au foil references together with their respective backgrounds (dashed lines). Background intensity increases with Au deposition duration. The background of the treated samples is higher when compared to the untreated sample having the corresponding Au deposition duration. This observation may be attributed to the greater island heights, which are associated with increased inelastic scattering when the deposition time is extended for the treated samples. In addition, it can be observed that the peak intensities of the untreated samples are closer to the those of the Au reference than the peak intensities of the treated samples are; this reflects that the coverage of the untreated samples is higher than that of the treated samples.

[0052] From the quantitative analyses of these spectra, the values of coverage c and height h have been obtained for each sample according to the Height and Surface Coverage Determination Method. Results are plotted in Fig. 3.

[0053] Fig. 3(a) depicts the amount of substance (AOS) vs. deposition time, defined as AOS = c x h, which represents the Au volume in terms of ‘nm of full surface coverage’. It is clearly seen that treated samples show a higher linear slope than the untreated samples, which reflects a higher deposition rate at longer deposition times. Fig. 3(b) illustrates the variation of surface height and coverage depending on the amount of Au deposited for both surfaces. Additionally, the series of dashed curves represent isolines for a fixed deposit (AOS) [8], A point situated on a dashed line (e.g.,2 nm) indicates that the amount of material is equivalent to a film of 2 nm thickness with 100% coverage, although the distribution could be different (e.g., 4 nm height with 50% coverage or 8 mm height with 25% coverage). Au deposited on treated surfaces has a tendency to grow more vertically (i.e. , perpendicular to the substrate), which is reflected by the higher values of h. In contrast, Au growth on untreated surfaces shows lower values of h and higher values of c, indicating a growth more ‘horizontal’, i.e., parallel to the substrate.Characterization of Samples using AFM

[0054] Atomic force microscopy (AFM) is a technique of surface analysis used to obtain images of micro / nanostructured films with resolutions of the order of fractions of nanometres. The basic working principle of AFM relies on a flexible cantilever with an integrated sharp tip (radius of around 10 to 50 nm) that scans the surface by means of an xyz scanner. AFM has many operation modes, and the tips can be operated and functionalized in several ways. Nevertheless, the typical outputs of this technique are topographical maps of the sample under analysis, i.e., sets of (x, y, z) triplets that indicate the height (z) of all the points (x, y) of a surface.

[0055] AFM images of samples have been analysed in the framework of the so-called dynamic scaling theory, which has been successfully used to describe the development of many systems exhibiting rough surfaces

[0010] , The underlying idea is that the surface remains similar to itself at different deposition times. The characteristics of a surface are constant regardless space dimension and deposition time if proper scaling factors are included. In other words, the characterizing parameters of the surface show invariance under a change of scale, which is a characteristic of fractals.

[0056] The surface morphology of the samples was analysed using the height-height correlation function (HHCF, H (r, t)) in order to estimate the growth parameters of the deposits: roughness (w), correlation length ( ) and Hurst parameter (a). The HHCF represents the average height difference squared between any pair of points, separated by a distance r over an entire AFM micrograph [9, 10]:H(r, t) = (| / i(r + r' , t) — h(r', t)|2) (Eq. 5) where h(r', t) is the surface height at a point r' and deposition time t, and “(□)” means averaging over all points r'. In practice, the averaging is limited to points on a scanningline (optionally followed by averaging over plural, e.g., parallel, lines). For a self-affine isotropic surface, several analytical forms have been proposed for the HHCF. Among the simplest ones is the one proposed by Sinha et al.

[0011] :where w, and a are the roughness, correlation length and roughness exponent (also called Hurst parameter). In Eq. 6, the deposition time dependency of H, w,and a has not been written (the deposition time t is fixed for a given sample). The interpretation of parameters w,and a is sketched Fig. 4. Roughness (w) represents the average deviation of heights of the sample with respect to the average height h. For similar values of w, a small value of the Hurst parameter (a « 0) implies a locally rougher surface, while a higher value (a « 1 ) is correlated with a smoother surface as seen in Fig. 4 (a). The correlation length,is the limit between both regimes, and it represents the lateral distance within the surface heights of any two points that are correlated, also known as the feature size (see Fig. 4 (b)). Combined, these three roughness parameters completely characterize a self-affine surface.

[0057] AFM micrographs of the samples were obtained using the AM-FM mode of the MFP-3D Infinity AFM from Asylum Research in tapping mode. The measurements were performed under standard ambient conditions (at room temperature and relative humidity of approximately 50%) using a standard cantilever holder for air operation. 200x200 nm2regions were imaged with a 256x256 pixel resolution at a scan rate of 3 Hz. The HHCF of each scanned horizontal line of the AFM micrograph was calculated, and it was averaged over all the lines of the AFM image. This approach is preferred to the HHCF of the whole AFM image. The reason is that each line is measured in a relatively short period of time, while the time difference between the measurement of several lines is larger, which may lead to drifts (e.g. thermal drifts). The HHCF calculations were performed using Gwyddion 2.61 . The parameters w,and a were obtained from fitting the experimental HHCF to Eq. 6 through a conventional approach of residual minimization (using Microsoft Excel®) including an appropriate weighting of the points located at low values of r, to ensure that the fitting goes through these data points and the fitting represents the experimental data. The value was taken as the Feature Size of the nanoparticles. These values were taken as averages of at leastthree AFM images acquired in different regions of each sample. The fittings of HHCF of one image of each sample are shown in Fig. 5. Fig. 5 (a) relates to the untreated surfaces, Fig. 5 (b) to the treated surfaces. For short values of r (r « ) the behaviour is a power law whose slope is controlled by a, and for long values of r (r » ) the value is a constant proportional to w.

[0058] The parameters w, and a obtained from fitting the experimental HHCF curves to Eq. 6 are depicted in Fig. 6. Fig. 6 (a) a shows the increasing nature of w with time from 0.23 to 0.84 nm. The roughness of the untreated samples is slightly higher than that of the treated samples for same deposition times. Fig. 6 (b) reveals an increase of with t for untreated samples, indicating a lateral growth of the Au nanoparticles. In contrast, is almost constant for treated samples. The parameters w and typically show a power law dependence

[0012] , as w ~ t and ~ t? , where / J and z are the so- called growth and dynamic exponents, respectively. These two exponents serve to describe the development of the surface with time

[0012] , The scaling exponents a, / Jand z can be related by z = for a self-affine surface

[0012] , This relationship was notverified with the analysed samples, indicating a more complex growth mechanism.

[0059] The growth parameter (3 was similar for both the untreated and the treated surfaces (similar slopes), with values of |3 = 0.82 ± 0.056 and [3 = 0.86 ± 0.099, respectively. Both values are relatively large, and they indicate that there are relevant non-local effects in the growth of the films (local effects may justify factors up to about 0.3).

[0060] In contrast, the behaviour of the dynamic parameter z is very different in untreated and treated samples: 1 / z = 0.32 ± 0.026 and 1 / z = -0.083 ± 0.041 , respectively. Experimentally, values between 0.13 and 0.85 have been reported, which indicates that the value of 1 / z is not universal and strongly depends on deposition conditions. The low value (near zero) of 1 / z for Au deposited on treated surfaces is particularly remarkable, since it indicates that the correlation length is almost constant from the beginning of the growth: in other words, from the beginning of deposition, the treated samples appear to have relatively large Au islands whose lateral size (feature size) does not change with deposition time.

[0061] The Hurst parameter, a, describes the undulation, and it lies between 0 and 1 , which represents a rougher and a smother local surface profile, respectively. Fig. 6 (c)shows a large increase in the Hurst parameter in the treated samples from 0.83 at 10 s to 0.99 at 80 s. In contrast, untreated samples show a much lower variation of a, from 0.95 at 10 s to 0.99 at 60 s. This indicates that treated surfaces show a ‘pointier’ growth at early stages of deposition. This observation is linked to the effect of Ar / H2 plasma treatment, which increases the traveling length of Au atoms on the alumina surface and promotes a locally rougher growth of gold compared to the untreated surface, where the a values are higher. However, as the deposition time increases, the value of a for treated surfaces approaches unity, in agreement with the values observed for the untreated surface. For additional characterization of the growth mode, the localslope, m ~ was plotted as a function of time, as illustrated in Fig. 6 (d). A highervariation on the slope with time is seen for treated surfaces. However, both set of samples change with time, a behaviour referred as non-stationary growth

[0010] , Additionally, an up-shift is seen for H(r) with increasing the deposition time, which indicates an increase in the local slope confirming the nonstationary growth [9, 10, 13],Feature Size Determination Method

[0062] A sample of a substrate comprising an oxide film and metallic nanoparticles on the surface of the oxide film is subjected to AFM so as to generate an AFM micrograph. The AFM micrograph is preferably carried out using the tapping mode of the atomic force microscope. The measurements may be performed under standard ambient conditions (at room temperature and relative humidity of approximately 50%) using a standard cantilever holder for air operation. The AFM micrographs shall cover regions of 200x200 nm2and shall be imaged with a 256x256 pixel resolution. At least three AFM micrographs of different regions of each sample have to be taken.

[0063] The HHCF of each scanned horizontal line of the AFM micrograph is calculated, and it is averaged over all the lines of the AFM image. These calculations may be performed using Gwyddion 2.61 .

[0064] The parameters w,and a are obtained from fitting the experimental HHCF to Eq. 6 through a conventional approach of residual minimization. Microsoft Excel® may be used for this. An appropriate weighting of the points located at low values of r shall be used, so as to ensure that the fitting goes through these data points and the fitting represents the experimental data. In other words, the data points located at low valuesof r (where the HHCF strongly depends on r) may have to be weighted higher compared to the data points located at high values of r (where the HHCF is essentially flat), in order to ascertain that the fitted curve correctly represents the data especially on the small scale. The value of is taken as the Feature Size of the nanoparticles. These values are taken as averages of at least three AFM images acquired in different regions of each sample.Optical Properties

[0065] The optical properties of nanocomposite films may be investigated by ultraviolet-visible (UV-Vis) spectroscopy. The working principle of this method involves the emission of a broad range of wavelengths of light, which then passes through a monochromator to select a specific wavelength. The resulting beam passes though the sample. When the beam interacts with the sample, it can be absorbed (la), reflected (lr) and the remaining light is transmitted (It). The intensity of the incident light (Io) can be expressed as the sum of the absorbed, reflected, and transmitted light intensities:

[0066] The absorbance, A, of a sample is defined as:A = - log? (Eq. 8). where T is lt / lo.

[0067] Samples US1 -US6 and samples TS1 -TS8 were analysed by UV-Vis spectroscopy. The absorbance spectra of the different samples are presented in Fig. 7. Fig. 7(a) displays the results for the untreated samples, Fig. 7 (b) those for the treated samples.

[0068] It is observed that the peak absorbance increases with deposition time. This is a consequence of a higher Au content, which leads to an increase in number and in size of absorption centres (Au NPs), which interact with the incoming light, causing an increase in the intensity of the absorbance peak

[0014] , In addition, for higher deposition times, the peak is shifted towards higher wavelengths because the particle shape changes progressively from a spherical towards a more worm-like structure and because of effects of size and local environment of the nanoparticles [15, 15], For untreated samples, a broad absorbance peak is for deposition times of 50 s and 60 s. This is associated with the transition from a morphology formed by (individual) Aunanoparticles and clusters to a continuous film (percolation). For treated samples, this transition occurs only at higher deposition times (> 80 s in Fig. 7 (b)). Therefore, the plasma treatment with H2 allows increasing the range of the plasmonic effect, widening the Au deposition range before percolation. As a result, treated samples showed a clear LSPR peak up to 80 s of deposition although the peak shape of TS8 reveals signs of percolation. Nevertheless, it is worth mentioning that the deposition rate was much higher for the treated than the untreated samples (cf. Fig. 3); in fact, the treated sample TS7 (deposition time of 70 s), which shows no percolation, carries about 2.5 times more Au than the untreated sample US5 (deposition time of 50 s), for which percolation is already observed. In other words, the plasma treatment in the H2-containing atmosphere allowed accommodating at least 2.5 more Au on the AI2O3 surface before the LSPR peak was lost.

[0069] Fig. 8 and Fig. 9 show the influence of different parameters on the efficiency of the plasma treatment.

[0070] Fig. 8 shows the absorption spectra of gold nanoparticles deposited for 60 s on a) amorphous AI2O3 (a- AI2O3) at room temperature and b) Y-AI2O3 at 450°C with / without Ar / H2 treatment. Fig. 8 (a) illustrates that the use of an Ar plasma treatment (no H2) induces a much lower effect than the use of Ar / H2 mixture. Fig. 8 (b) indicates that the Ar / H2 plasma treatment is effective on crystallized y-alumina but the effect is less than on amorphous alumina.

[0071] Fig. 9 shows the absorption spectra of gold nanoparticles deposited for 30 s on alumina surfaces treated in different Ar / H2 plasma conditions. Fig. 9 (a) relates to an H2 flux of 1 .5 seem, Fig. 9 (b) to an H2 flux of 3 seem. The durations of the plasma treatments were 10, 20 and 30 minutes. For comparison, Fig. 9 shows the absorption peaks of US2 and US3 (dashed curves). Fig. 9 indicates that the concentration of H2 in the atmosphere of the plasma treatment and the duration of the plasma treatment (between 10 and 30 min) do not seem to induce any major change. In other words, the sole presence of H2 is enough to induce the effect, the quantity and the duration are of less relevance.

[0072] In a separate experiment, it was shown that use of a pulsed-DC source to sputter Au instead of a conventional DC source did not show any influence on the optical properties of the deposited films.Areal Number Density Determination Method

[0073] A mean interparticle distance and particle density may be calculated from h, c (obtained with the Height and Surface Coverage Determination Method) and £ (obtained with the Feature Size Determination Method) through a geometrical model, which assumes identical nanoparticles with axis-symmetric 3D Gaussian shapes distributed in an hexagonal-packaged arrangement. The equation of a (circular symmetric) Gaussian function is:+(Eq. 9a) where (xo, yo) represents the location of the centre of the peak, A is the height, and cr represents its width.

[0074] By setting:A = h (Eq. 9b) and 2<J = (Eq. 9c) the mean interparticle (centre-to-centre) distance, d, may be calculated with: d2= c x V3 ( \Eq “. 10) /

[0075] The number density of particles per unit of surface (p) is calculated considering that there are three nanoparticles per hexagon (one in the centre and six in the vertices, each nanoparticle in a vertex being shared by three hexagons), as:Synthesis of Characterisation of Samples TS1-TS8 and US1-US6

[0076] Table 1 summarizes the different characteristics of the samples TS1 -TS8 and US1 -US6 obtained from XPS, AFM and optical characterisation as discussed above.Table 1In the table, t is the deposition time, c is the coverage, h the height, a the Hurst parameter,the correlation length (feature size), w the roughness, A the spectral position of the absorbance peak, A the absorbance, h / t; the aspect ratio, d the centre- to-centre interparticle distance and p the areal number density.

[0077] In general, SERS substrates subjected to an intermediary plasma treatment in an H2 atmosphere show much higher aspect ratio (h / S,), larger interparticle distance (d) and lower particle density (p), which explains the later onset of percolation and the higher ability to accommodate Au nanoparticles on the surface.Measurement of SERS effects

[0078] Selected A11-AI2O3 coatings were tested with Rhodamine 6G (R6G) for Surface-Enhanced Raman Spectroscopy (SERS) applications. On each sample, surface regions with and without Au nanoparticles were created in direct vicinity to each other by deliberately scraping the nanoparticles. The zones devoid of Au NP serve for comparison.

[0079] Untreated and treated samples were analysed by Raman Spectroscopy, using R6G. Fig. 10 (a) shows that untreated samples have lower peak intensity than treated samples. R6G in a concentration of 10’5mol / L was used for these measurements. This observation validates that the utilization of H2 plasma pre-treatment enhances R6G detection by inducing alterations in Au nanoparticle growth.

[0080] Another measurement was made on treated samples with different Au deposition times: 30, 50, and 70 s, using R6G in a concentration of 10-3mol / L. Fig. 10 (b) illustrates the Raman spectra for the three different samples. In the AI2O3 zone (freed from Au nanoparticles), spectral peaks corresponding to the silicon peaks of the substrate at 301 , 520, and between 935 and 990 cm-1are observed

[0017] , In contrast,the spectrum of R6G can be clearly observed in the regions of the samples covered with Au nanoparticles, which confirms the SERS capabilities of these samples. The results reveal that the signal experiences enhancement as the deposition time of Au increases, with the highest enhancement observed for the sample having undergone Au deposition for 70 s.

[0081] SERS measurements were conducted in several regions of the samples, and the measured spectra were found to be in nearly perfect agreement, which confirms excellent reproducibility.

[0082] Another important property of SERS substrates is shelf-life. The SERS experiments were carried out more than two years after the production of the samples. The samples had been stored in boxes under air, without any additional protection or precaution. This indicates that treated samples show a strong SERS effect even after long storage time, which is an advantage for potential commercial applications. This may be due to the high chemical stability of the samples (comprising alumina as the oxide film and Au nanoparticles) against oxidation and other processes of environmental degradation.

[0083] While specific embodiments have been described herein in detail, those skilled in the art will appreciate that various modifications and alternatives to those details could be developed in light of the overall teachings of the disclosure. Accordingly, the particular arrangements disclosed are meant to be illustrative only and not limiting as to the scope of the invention, which is to be given the full breadth of the appended claims and any and all equivalents thereof.List of References

[0001] X.X. Han, R.S. Rodriguez, C.L. Haynes, Y. Ozaki, B. Zhao, Surface-enhanced Raman spectroscopy, Nature Reviews Methods Primers. 1 (2022) 87. https: / / doi.Org / 10.1038 / s43586-021 -00083-6.[2] S. Tougaard, Universality classes of inelastic electron scattering crosssections, Surface and Interface Analysis. 25 (1997) 137-154. https: / / doi.org / 10.1002 / (SICI)1096-9918(199703)25:3<137::AID-SIA230>3.0.CO;2-L.[3] S. Tougaard, Quantitative analysis of the inelastic background in surface electron spectroscopy, Surface and Interface Analysis. 11 (1988) 453-472. https: / / doi.Org / 10.1002 / sia.740110902.[4] S. Tougaard, Energy loss in XPS: Fundamental processes and applications for quantification, non-destructive depth profiling and 3D imaging, J Electron Spectres Relat Phenomena. 178-179 (2010) 128-153. https: / / doi.Org / 10.1016 / j.elspec.2009.08.005.[5] S. Tanuma, C.J. Powell, D.R. Penn, Calculations of electron inelastic mean free paths. V. Data for 14 organic compounds over the 50-2000 eV range, Surface and Interface Analysis. 21 (1994) 165-176. https: / / doi.org / 10.1002 / sia.740210302.[6] H. Shinotsuka, S. Tanuma, C.J. Powell, D.R. Penn, Calculations of electron inelastic mean free paths. X. Data for 41 elemental solids over the 50eV to 200keV range with the relativistic full Penn algorithm, Surface and Interface Analysis. 47 (2015) 871-888. https: / / doi.Org / 10.1002 / sia.5789.[7] S. Hajati, S. Tougaard, What nano-physical properties can be determined by analysis of elastic peak accompanied by its inelastic background tail in XPS and AES spectra?, 2006.[8] C. Mansilla, F. Gracia, A.I. Martin-Concepcion, J.P. Espinos, J.P. Holgado, F. Yubero, A.R. Gonzalez-Elipe, Study of the first nucleation steps of thin films by XPS inelastic peak shape analysis, Surface and Interface Analysis. 39 (2007) 331-336. https: / / doi.Org / 10.1002 / sia.2509.[9] R. Kesarwani, P.P. Dey, A. Khare, Correlation between surface scaling behavior and surface plasmon resonance properties of semitransparent nanostructured Cu thin films deposited via PLD, RSC Adv. 9 (2019) 7967-7974. https: / / doi.Org / 10.1039 / C9RA00194H.

[0010] Matthew Pelliccione, Toh-Ming Lu, Evolution of Thin Film Morphology, Springer New York, New York, NY, 2008. https: / / doi.org / 10.1007 / 978-0-387-75109-2.

[0011] S.K. Sinha, E.B. Sirota, S. Garoff, H.B. Stanley, X-ray and neutron scattering from rough surfaces, Phys Rev B. 38 (1988) 2297-2311. https: / / doi.Org / 10.1103 / PhysRevB.38.2297.

[0012] Matthew. Pelliccione, T.-M. (Toh-M. Lu, Evolution of thin film morphology : modeling and simulations, Springer, 2008.

[0013] J.H. Jeffries, J.K. Zuo, M.M. Craig, Instability of Kinetic Roughening in Sputter- Deposition Growth of Pt on Glass, Phys Rev Lett. 76 (1996) 4931-4934. https: / / doi.Org / 10.1103 / PhysRevLett.76.4931 .

[0014] N.M.R. de Figueiredo, Nanocomposite coatings based on metal nanoparticles exhibiting Surface Plasmon Resonance deposited by sputtering techniques, 2018. http: / / hdl.handle.net / 10316 / 79523.

[0015] S. Christke, C. Katzer, V. Grosse, F. Schmidl, G. Schmidl, W. Fritzsche, J. Petschulat, T. Pertsch, M. Rettenmayr, Optical resonances of self-organized monocrystalline Au nanoparticles embedded in SrTiO_3 matrix, Opt Mater Express. 1 (2011 ) 890. https: / / doi.Org / 10.1364 / ome.1 .000890.

[0016] N.M. Figueiredo, T. Kubart, J. A. Sanchez-Garcia, R. Escobar Galindo, A. Climent-Font, A. Cavaleiro, Optical properties and refractive index sensitivity of reactive sputtered oxide coatings with embedded Au clusters, J Appl Phys. 115 (2014) 0-13. https: / / doi.Org / 10.1063 / 1 .4861136.

[0017] W.J. Lee, Y.H. Chang, Growth without postannealing of monoclinic VO2 thin film by atomic layer deposition using VCI4 as precursor, Coatings. 8 (2018). https: / / doi.Org / 10.3390 / coatings8120431 .

Claims

Claims1. A method for the fabrication of a nanocomposite film, e.g., a SERS or LSPR substrate, comprising depositing an oxide film on a substrate, e.g., by physical vapour deposition; depositing metallic nanoparticles on the oxide film; characterised in that, prior to the deposition of the metallic nanoparticles, the oxide film is subjected to a plasma surface treatment in a reducing atmosphere.

2. The method as claimed in claim 1 , wherein the oxide film is deposited by sputtering, e.g., by HiPIMS.

3. The method as claimed in claim 1 , wherein the oxide film is deposited by reactive sputtering in an atmosphere including O2.

4. The method as claimed in any one of claims 1 to 3, wherein the oxide of the oxide film comprises at least one of AI2O3, CuO, WO3, and SiO2.

5. The method as claimed in any one of claims 1 to 4, wherein the metal of the metallic nanoparticles comprises at least one of Au, Ag, Cu, Pt and Pd.

6. The method as claimed in any one of claims 1 to 5, wherein the reducing atmosphere of the plasma surface treatment consists of or includes H2.

7. The method as claimed in claim 6, wherein the atmosphere of the plasma surface treatment comprises a mixture of H2 and a gas selected from He, Ar, Ne, Kr, Xe.

8. The method as claimed in any one of claims 1 to 7, wherein the plasma of the plasma surface treatment is generated with a microwave source or an RF plasma source, e.g., an ICP source or a CCP source.

9. The method as claimed in any one of claims 1 to 8, wherein the oxide of the oxide film comprises AI2O3.

10. The method as claimed in any one of claims 1 to 8, wherein the oxide of the oxide film comprises AI2O3, the metal of the metallic nanoparticles Au and the atmosphere of the plasma surface treatment comprises a mixture of H2 and Ar.

11. The method as claimed in any one of claims 1 to 10, wherein the deposition of the metallic nanoparticles on the oxide film comprises the formation of metallicnanoparticles having an aspect ratio h / t; (height / feature size) in the range from 0.65 to 1.50.

12. The method as claimed in any one of claims 1 to 11 , wherein the deposition of the metallic nanoparticles on the oxide film comprises the formation of metallic nanoparticles with an areal number density less than 6000 pm-2.

13. The method as claimed in any one of claims 1 to 12, wherein the deposition of the metallic nanoparticles on the oxide film comprises the formation of a surface coverage of the metallic nanoparticles on the oxide film lies in the range from 30% to 50%.

14. The method as claimed in any one of claims 1 to 13, wherein the deposition of the metallic nanoparticles on the oxide film comprises the formation of a nonpercolated network of the metallic nanoparticles on the oxide film.

15. A substrate, e.g., a SERS or LSPR substrate, comprising an oxide film on the surface of the substrate and metallic nanoparticles deposited on the oxide film, wherein the metallic nanoparticles are characterized by an aspect ratio h / t; (height / feature size) in the range from 0.65 to 1 .50.

16. The substrate as claimed in claim 15, wherein the metallic nanoparticles are characterized by an areal number density less than 6000 pm-2.

17. The substrate as claimed in claim 15 or 16, wherein the surface coverage of the metallic nanoparticles on the oxide film lies in the range from 30% to 50%.

18. The substrate as claimed in any one of claim 15 to 17, wherein the metallic nanoparticles form a non-percolated network on the oxide film.

19. The substrate as claimed in any one of claim 15 to 18, wherein the oxide film comprises or consist of at least one of AI2O3, WO3, CuO, and SiCh.

20. The substrate as claimed in any one of claim 15 to 19, wherein the metal of the metallic nanoparticles comprises at least one of Au, Ag, Cu, Pt and Pd.21 . The substrate as claimed in any one of claim 15 to 20, wherein the oxide of the oxide film comprises AI2O3, and wherein the metal of the metallic nanoparticles comprises Au.

Images