Align MWCNT/GNPs/PDMS based nanocomposite dry ECG electrodes for ECG recordings

2.1. Materials

The materials used for developing ECG electrodes consist of AMWCNT and GNP mixed with different weight ratios (5 %, 7 %, 9 %) in the PDMS matrix, which is given in Table 1.

AMWCNTs were obtained from Nanografi (Ankara, TURKEY), which has > 95 % purity. Nanotubes in terms of diameters 10-20 nm and length is between 1-2 µm. In addition, it has a 2.5 g/mL density at room temperature. GNPs were also purchased from Nanografi, it has a purity of over 99.9 %. GNP’s thickness is around 5 nm and its diameter is 18 nm. Sylgard 184 Silicone Elastomer (PDMS) is the polymer used because of its impressive characteristics like biocompatibility, non-toxicity, and flexibility and purchased from Dow Corning, Michigan, USA.

Fig. 2A schematic illustration of four different modifications for CNTs or GNPs interacting with polymer matrix [19]

The ideal dispersion of material inside the composite is related to the application, and how they were dispersed showed Fig. 3 and isopropyl alcohol (IPA) was used for dispersing CNTs and GNPs in PDMS. Even though, CNT-GNP nanocomposites might not be as tough or solid as a reinforced composite such as carbon fiber laminates. However, the multifunctionality of CNTs and graphene is still becoming game-changers. In addition to the various challenges of dispersion, viscosity control, sizing, and alignment of these materials.

2.2. Fabrication of ECG electrodes

Multi-walled carbon nanotubes and graphene nanoplatelets were weighed the desired ratio. Then, 3 different percentages of AMWCNT/GNP according to weight of PDMS were added to the beakers. These percentages are; 5 %, 7 %, 9 %. The calculations were made considering the 6:1 ratio. According to the ratios and concentration, the mixtures were named as PDMS5, PDMS7, and PDMS9. AMWCNT and GNP were first dispersed in 50 ml of isopropyl alcohol (IPA) using magnetic stirrer for 10 minutes. The solution was kept in an ultrasonic bath at 40 kHz for 30 minutes. After uniform dispersing was obtained, of the PDMS A form with 10 gr was mixed with the solution followed by mechanical stirring for 20 minutes. The composite solution was then placed on a hot plate with magnetic stirring for exsiccating of IPA and overnight waited at 100 degrees. After that, a PDMS B form (crosslinking agent) was added in 10:1 into the composite and degassed completely in a can be used vacuum. After the exsiccation of air in the mixture, nanocomposites were immediately molded on already prepared molds, these steps are shown in more detail in the following Fig. 3. AMWCNT/GNP/PDMS hybrid nanocomposites were cured in prepared steel molds at 120 °C for 30 minutes.

Fig. 3Schematics of composite fabrication for ECG electrodes

3.1. Conductivity

PDMS5, PDMS7 and, PDMS9 samples were cut into 1×1 cm² rectangular pieces. The thickness of each hybrid nanocomposite was measured with a digital micrometer. Thickness measurements of each hybrid sample were made from 3 different points previously determined on the surface and made averages. Thickness measurements of each hybrid sample are shown in Table 2.

The electrical conductivity measurements of the samples were calculated from the plate resistance ($R$) values obtained by using flux, and the siemens measurement (S/m) unit was used. Here, firstly, as a result of the resistance (Ohm.m) values obtained, the transition to resistivity ($\rho$) is made by using Eq. (1). Then, Eq. (2) was applied to the resistivity values found:

The correlation and variation between nanocomposite electrodes are shown in the Fig. 4. It was observed that the electrical conductivity of the nanocomposite structure increased as the weight of the carbon nanotube and graphene placed in the mixture increased. Although this change is not linear, it increases exponentially in the percolation threshold region. That is quite low for aligned carbon nanotubes graphene nanoplates nanocomposites.

Fig. 4Conductivity of the produced electrodes is in the form of siemens

3.2. Fourier transform infrared spectroscopy (FTIR)

FTIR Spectrum100 (Perkin Elmer, Waltham, MA) was used. The frequency range was 650-4000 cm^-1 wavenumber in the transmittance mode and FTIR is one of the important characterization techniques for identifying functional groups in AMWCNT/GNP/PDMS.

Structural properties of the different ratios at AMWCNT/GNP/PDMS were measured by FTIR spectroscopy in the frequency range of 650-4000 cm^-1 (Fig. 5). The FTIR spectrum of AMWCNT/GNP/PDMS was characterized by a symmetrical stretching vibration of -CH₃ at 2962 cm⁻¹, deformation vibration of -CH₃ in Si-CH₃ at 1257 cm^-1, Si-O-Si asymmetrical deformation at 1005 cm^-1, C-O-S stretching at 911 cm^-1. The absorption band appeared at 863,789 cm^-1, respectively, representing -CH₃ symmetrical rocking. The peaks correspond to Si-C asymmetrical stretching, Si-C symmetrical stretching, and Si-CH₃ symmetrical rocking in nanocomposite were observed at 843,758, and 687 cm^-1, respectively. As can be seen from the results obtained, mixing AMWCNT/GNP to PDMS at increasing rates changed the intensities of the PDMS’s bands and led to the emergence of new bands in the spectrum that would be attributed to AMWCNT/GNP. This indicates that there is an interaction between PDMS and AMWCNT/GNP.

3.3. ECG measurement setup

From the results obtained for conductivity, it can be said that the nanocomposite of PDMS with align-MWCNTs proved to be a better choice and preferable for ECG monitoring. The ECG measurement system was obtained through an Arduino Uno board coupling with an AD8232 cardiac sensor board. Therefore, the cardiac cycles of producing flexible electrodes have been obtained using Arduino software.

An ECG measurement was taken from the chest of a subject for 15 seconds using AMWCNT/GNP/PDMS electrode and as a control Ag/AgCl dry electrode and compared.

Fig. 5FTIR spectra of AMWCNT/GNP/PDMS at differential ratios

3.4. Heart cycles with Arduino uno board as a microcontroller

The ECG signal quality, while the body is fixed, was then analyzed and compared between the wet Ag/AgC1 electrode and the printed dry ECG electrodes. The cardiac cycles of produce flexible electrodes have been obtained by used software in Arduino and have been shown that comparative with dry-electrode in Fig. 6. When the ECG signal of AMWCNT/GNP/PDMS and Ag/AgCl electrodes are compared at the same time, it seems that there is no significant difference between the two signals. When the peaks are examined considering the waves occurring in the cardiac cycle, it is seen that P, QRS and T waves can be clearly selected. The waveform of the ECG signal from the AMWCNT/GNP/PDMS electrode almost exactly matches that from the Ag/AgCl electrode.

Fig. 6Schematics of MWCNT/GNP/PDMS heart cycles in ECG electrodes

In line with the data, all cardiac cycle waves are seen in the ECG measurements taken. It is seen that the PDMS9 (Fig. 6) electrodes transmit the cleanest signal. However, studies on these electrodes should continue in the future and the most accurate result can be achieved by using signal processing. The data of all cardiac cycle waves should be examined separately as a long term.