*2.2. Materials Characterization*

Four of the most used fabrics (organic cotton, jeans, viscose and lycra) were tested in order to take into consideration a broad selection of daily use textiles (Figure 1a–d) [38,39]. In the pictures, the di fferent porosity of the fabrics can be seen; these air voids have an impact on the variation of the e ffective permittivity. Some studies have shown that there is a linear variation of the relative permittivity depending on the infill percentage [40]. An increase in the density of porous within the fabric substrate implies more air voids trapped. The dielectric constant of the air (<sup>ε</sup>r = 1) is lower than the fabrics, lowering the total dielectric constant and increasing the dissipation factor value (tan δ). The dissipation factor (DF) (often known as loss tangent, tan δ) is a ratio of the loss index (ε") and the relative permittivity (ε )

$$\text{LPF} = \tan \delta = \varepsilon'' / \varepsilon' = 1/\mathbf{Q} \tag{5}$$

where Q is the quality factor and it describes how underdamped a resonator is.

**Figure 1.** Microscope images for textiles substrates: (**a**) Cotton; (**b**) Jeans; (**c**) Viscose and (**d**) Lycra.

From the microscope images (Figure 1a–d), we extracted the morphology of the four textile substrates. The first three fabrics (organic cotton, jeans and viscose) were woven fabrics with multiple fibres crossing each other at di fferent angles to form the grain, while the last one (lycra) was knitted, it was made up of a single yarn, looped continuously to produce a braided look. The two major fabric types were considered in this investigation.

For the second part of the research, we will focus on using organic cotton due to its presence in daily clothing. The values of dielectric constant and dissipation factor for the four textile substrates measured using the resonant method of cavity perturbation are given in Table 1 below.



### *2.3. Antenna Design and Fabrication*

The antenna design falls into a modelled low-profile microstrip patch antenna that was slightly adapted for each of the three frequencies of operation considered in this case of study: 2.45 GHz (A), 9.45 GHz (B) and 38 GHz (C).

### 2.3.1. Case A at 2.45 GHz

The first case was an inset feed microstrip patch antenna (Figure 2a), designed to operate around 2.45 GHz (industrial, scientific and medical band, ISM) using the Equations (2)–(4). Figure 2a represents

the layout of the proposed antenna for the four textile substrates used during the numerical analysis. The corresponding parameters dimensions in mm are illustrated in Figure 2b.

**Figure 2.** (**a**) Modelled textile antenna (**b**) Antenna parameter dimensions.

Regarding materials, adhesive copper tape (with conductivity σ = 5.8 × 10<sup>7</sup> <sup>S</sup>·m<sup>−</sup>1) was used to construct both the radiation patch and the full ground plane. In order to minimize errors both pieces were cut using a Graphtec Craft-Robo CC300 from Materials Engineering Laboratory at Queen Mary University of London (QMUL). These two pieces were manually attached with a thin layer of acrylic pressure-sensitive conductive adhesive with good heat resistance to the low-cost textile substrates. The impact of the thin layer of adhesive can be ignored for the purpose of this study. Finally, a standard SMA 50 Ohm (Ω) connector was soldered to both, the edge-fed of the antenna and the background plane. Final prototypes are depicted below (Figure 3a–e).

**Figure 3.** (**a**) Cotton top-view; (**b**) Cotton bottom-view; (**c**) Jeans; (**d**) Viscose; (**e**) Lycra; (**f**) Cotton Case B—9.45 GHz and (**g**) Cotton Case C—38 GHz.

### 2.3.2. Case B at 9.5 GHz

The second case of study was carried out in order to increase the sensitivity of the system. We kept the same simple approach of using a microstrip patch antenna. In this case, a microstrip feed was designed and fabricated to work at higher frequencies, around 9.5 GHz (Figure 3f) [13,41]. We moved the operational frequency band from the S-band (ISM, 2.45 GHz) towards the X-band (8–12 GHz). Same Equations (2)–(4) were used for designing the antenna, in which the antenna patch dimensions are 13.8 mm × 9.8 mm (width × length). These reduced dimensions imply resonating at higher in frequencies, due to the fact that the antenna size is related with the wavelength (λ) of operation, which in turn is inversely proportional to the resonant frequency

$$
\lambda = \mathbf{c}/\mathbf{f}\_{\mathbf{f}} \tag{6}
$$

For fabrication, the same process and techniques were followed and as substrate material, the same organic cotton was used. To interface with the vector network analyser (VNA) during the test campaigns the same connector was soldered.

### 2.3.3. Case C at 38 GHz

For the final case, a new antenna was designed and fabricated, following the same rationale of previous Section 2.3.2. The design proposed in [42] has been used for this frequency, optimizing the dimensions of the radiating square patch 2.7 mm × 2.8 mm (width × length) and the position of the stub through the feeding line (Figure 3g). An SSMA 2.92 mm edge-launch connector working at 38 GHz was soldered. This connector mates with the popular SMA connections from most of the laboratory pieces of equipment. In addition, this fabric (organic cotton) antenna was designed and fabricated to work at millimeter-waves (mmW), with the potential to be used within the range of the emerging 5G technology.
