1. Introduction
The use of coatings for aesthetic purposes is a widely studied field with many applications. These coatings can be applied through various techniques, such as painting [
1,
2], electrochemical deposition [
3,
4,
5], chemical vapor deposition (CVD) [
6], and physical vapor deposition (PVD) [
7,
8,
9,
10], among others. PVD techniques are a commonly employed choice across different industries for aesthetic enhancement, such as metalizing components, especially in the automotive sector [
11,
12], or applying them to building glazings [
13].
PVD techniques, particularly magnetron sputtering, enable the deposition of thin film coatings composed of a wide array of materials, while providing precise control over the thickness of these layers. This capability allows for the fabrication of interference optical coatings, achieving an aesthetic appearance and optical properties that are unattainable with simple paints [
14,
15]. Moreover, the spectral transmittance of the coatings can be meticulously controlled. Interference optical coatings enhance the optical properties of surfaces and are extensively applied in the production of antireflective coatings, optical filters, low-emissivity glasses, and more [
16,
17,
18,
19].
The calculation used to determine the optical properties of such coatings is well established, with several models available for this purpose [
18,
20,
21]. Open access programs also exist, facilitating both the calculation of these properties and the determination of optimal layer thicknesses based on a given target spectrum [
22].
The fabrication of dichroic optical filters, employing dielectric materials, allows for the attainment of various spectral curves, with the constraint that they exhibit no optical absorption. Consequently, the sum of the transmittance and reflectance values equals one for all wavelengths. However, when the goal is to develop coatings with optical absorption, aiming to achieve relatively independent transmittance and reflectance, the use of materials with optical absorption is necessary [
23,
24]. In other words, these are materials with an imaginary component in their refractive index. This can be easily accomplished through the utilization of metallic materials; nevertheless, the high electrical conductivity of these materials results in the high reflectance (and absorption) of electromagnetic waves in the radiofrequency range [
25,
26].
A parameter tightly linked to the absorption, transmission, and reflection of radiofrequency in thin films is sheet resistance. In [
27], expressions for transmitted amplitudes are calculated based on the various physical parameters of thin films, such as thickness and conductivity. From the study of these expressions, it was deduced that, to achieve high-level radiofrequency transmission, it is necessary for R
s to be significantly greater than Z
0, where R
s represents sheet resistance and Z
0 is the impedance of a vacuum (377 Ω). Conversely, when R
s is much lower than Z
0, the coating will exhibit high radiofrequency attenuation.
In the current landscape, the significance of electromagnetic waves in the radiofrequency range has surged, given the widespread use of mobile phones that operate across various radiofrequency bands, spanning from 0.7 to 3.6 GHz. Consequently, when considering coatings for construction applications, it is crucial for them to exhibit less attenuation within these radiofrequency bands. This consideration has gained particular prominence in the domain of low-emissivity coatings, prompting the development of solutions that make use of frequency-selective surfaces (FSS) [
28,
29]. In the automotive industry, the development of decorative coatings with minimal radiofrequency attenuation is also a noteworthy pursuit, driven by the current utilization of radar systems in vehicles.
This study focuses on the development of coatings with various aesthetic appearances using germanium as a material with optical absorption. Germanium, being a semiconductor material with a bandgap energy of 0.66 eV, exhibits absorption in the visible spectrum, but has less conductivity compared to that of metals. This characteristic enables the creation of coatings with low radiofrequency attenuation, while achieving diverse aesthetic effects, such as black coatings, metallic gray coatings, or coatings in different colors [
30,
31]. Moreover, these coatings may have semi-transparency in the visible spectrum, and multilayer structures are devised to modulate the visible transmittance factor of the coating.
2. Materials and Methods
The samples were deposited using DC pulsed magnetron sputtering in a semi-industrial facility, where the samples moved linearly facing the target. The target sizes were 600 mm × 100 mm, and the large size of the materials combined with their dynamic deposition allowed for the production of coatings with homogeneous and uniform thickness on substrates measuring up to 520 mm × 300 mm. ClearTrans ceramic glass substrates from Schott were used, which were cleaned with ACEDET detergent prior to the deposition process. Additionally, a treatment to clean and activate the surface was conducted using an Ar+ ion beam generated by introducing a flow of 50 sccm (standard cubic centimeters per minute) of Ar into the process chamber and applying an accelerating voltage of 2 kV.
The materials employed included 99.99% pure germanium (Ge) and a silicon (90%)/aluminum (10%) alloy with a purity of 99.99%. The deposition process took place under a base pressure below 10−4 Pa and a working pressure on the order of 10−1 Pa. The deposition of Ge layers involved introducing a flow of 200 sccm of Argon into the chamber (corresponding to a pressure on the order of 10−1 Pa) and applying a power of 700 W (1.17 W/cm2). The deposition of SiAlNx layers was carried out using reactive sputtering, introducing a flow of 100 sccm of Argon and 100 sccm of N2. A power of 2.5 kW (4.17 W/cm2) was applied during this process.
The deposition rate and layer thickness were calibrated using a Dektak XT (Bruker, Billerica, MA, USA) mechanical profilometer. The layers of each material (Ge and SiAlNx) were individually deposited under the same deposition conditions as those of the decorative coatings. However, a slower substrate holder speed was employed, resulting in a longer sample deposition time, aimed at achieving layers of approximately 100 nm thickness and enhancing the thickness calibration precision. Subsequently, for the deposition of the decorative coatings, the substrate holder speed was adjusted to achieve the desired thickness for each coating layer. This adjustment followed an inverse proportionality rule, wherein a faster substrate holder speed corresponded to a shorter deposition time and thinner deposited layer.
Spectrophotometric measurements were conducted using a homemade UV-VIS-IR spectrophotometer, which measures both transmittance and reflectance at an incidence angle of 8° across the range from 300 to 2500 nm.
The complex refractive indices of the materials were calculated from the spectrophotometric measurements of transmittance and reflectance (from both sides) of the samples with only a deposited layer of approximately 30–50 nm. To determine the refractive indices of this layer, its thickness was measured, and a simple model of a layer on a substrate with a known thickness and refractive index was built. The system was numerically simulated, calculating each pair of values (
n is the real part of the refractive index, and
κ is the imaginary part), and an attempt was made to minimize the defined merit function
χ as follows:
where
Tcalc,
Rccalc, and
Rcalc represent the theoretically calculated transmittance, coating-side reflectance, and substrate-side reflectance, respectively, for each value of
n and
κ. This calculation was performed using a single-layer model of Air/Substrate/Film/Air utilizing a known substrate (its refractive indices) and taking into account the generated interferences. Meanwhile,
Texp,
Rcexp, and
Rexp denote the experimentally measured values for the characterized sample. Merit function minimization was performed at each wavelength
λ, thereby adjusting the values of the complex refractive index for all the spectrophotometrically measured wavelengths. From the imaginary part of the refractive index, the absorption coefficient
α was calculated as follows:
The absorption coefficient of germanium calculated for different wavelengths was used to determine the bandgap energy through the Tauc plot [
32], employing the following expression:
where
r represents a value depending on the type of transition, with
r = 1/2 denoting indirect transitions. This method takes into account multiple reflections, resulting in more accurate results [
33].
The optical properties were theoretically adjusted using proprietary software based on the continuity of tangential components of the electric and magnetic fields at the interfaces of the multilayer structure [
20].
The transmittance and reflectance spectra measured with a spectrophotometer were used to calculate the values of visible transmission
TVIS and visible reflection
RVIS using the following expressions:
where
T(λ) represents spectral transmittance,
R(λ) represents spectral reflectance,
λ represents the wavelength,
D65 represents the D65 illuminant defined by the International Commission on Illumination, and
V(λ) represents the sensitivity of the human eye. This calculation was performed following the EN 410 standard [
34]. For color calculation, CIELab color coordinates were utilized, distinguishing between the color coordinates in transmission
LABTrans and reflection
LABReflec.
Structural and morphological characterizations were performed using XRD with an RIGAKU device model D/max 2500 (Tokio, Japan) and using Field-Emission Scanning Electron Microscopy (FESEM) with a Carl Zeiss MERLIN model (Oberkochen, Germany).
The experimental setup for RF measurements (
Figure 1) used two directional Vivaldi antennas (TSA600, RFSPACE, Atlanta, GA, USA) operating within the 600–6000 MHz frequency range and placed in an anechoic chamber. Signal generation and measurements were carried out using a vector network analyzer (picoVNA 106, Pico Technology, Cambridge, UK).
For the characterization of the different glass samples, a first measurement was made without placing anything between the antennas, which was taken as a reference. This was used to discount the effect of the cables, antennas, or secondary reflections on the anechoic chamber. Then, the glass sample was placed between the antennas, and a second measurement was taken. Attenuation was calculated as the difference between both these measurements.