Biomimetic Synthesized Conductive Copolymer EDOT-Pyrrole Electrodes for Electrocardiogram Recording in Humans

We report on electrodes fabricated with EDOT-Pyrrole copolymer through electrophoretic deposition and used for recording and sensing bio-electrical signals. We measured the electrical properties of the copolymer deposited on a stainless-steel substrate, and we performed Cyclic Voltammetry (CV) and Scanning Electron Microscopy (SEM) studies to characterize the morphological properties and copolymer distribution on the metal surface. We found that electrodes fabricated with EDOT-Pyrrole copolymer exhibit a high sig-nal-to-noise ratio as well as an accurate and stable conductivity compared with other commonly used electroconductive polymers. Stainless-steel-coated EDOT-Pyrrole electrodes are suitable to record electrocardiograms in humans with high resolution comparable to standard silver-electrodes.

clinical tools which provides physicians the preliminary diagnosis of cardiac complaints. The electrodes are a critical part of the electrocardiogram recording (ECG) and their evolution in the past century has been dramatic, from cylindrical electrolyte electrodes to the wearable electrodes commonly used in the last few years. Typically, an ECG electrode is made of silver/silver chloride (Ag/ AgCl), due to low impedance, high conductivity, and electrical stability in its system. However, the shape and size of this kind of electrode are limited due to biocompatibility effects.
To overcome the biocompatibility effects, recent advances in electrode materials have been made, in order to achieve comfortable, flexible and biocompatible electrodes. A wide variety of materials have been employed, such as carbon nanomaterials [1], carbon nanotubes [2], metallic nanoparticles [3], inorganic semiconductors [4], and electroconductive polymers. Conductive polymers are very promising for electrode materials because they are soft, electrically conductive, low cost, easily synthesized, and biocompatible with human tissues [5] [6].
One of the most fundamental characteristics of bioelectrodes is their biocompatibility. PEDOT:PSS, as well as other conductive polymers that are usually biocompatible. However, it has been found that the polymer synthesis method may affect the biocompatibility of the material as the common chemical oxidative polymerization pathway produces a broad of toxic residues that are difficult to remove after synthesis. An alternative is an electrochemical synthesis, however through this technique the size and shape of the fabricated electrodes is limited, and it is not possible to generate high mass loading. In the last few years, a new synthesis approach emerged based on the enzymatic activity for catalysis of non-biological reactions like the conductive polymer synthesis [14]. This method produces polymers with much fewer toxic residues. However, the high susceptibility of the enzymes under the typical low pH conditions demanded by that biomimetic synthesis allows to mimicry enzyme catalyst function using organic molecules to bypass the enzyme's fragility, which results in a robust pathway for conductive polymer synthesis compared to enzymatic synthesis [15].
The Hematin molecule has been used by our research team for the biomimetic synthesis of polyaniline and poly-pyrrole (PPY) [15]. The second major problem concerning conductive polymer bioelectrodes is the manufacturing method.
Several approaches have been proposed in the last 30 years, including, spincoating, inkjet printing, electrochemical deposition, 3D printing, in situ chemical synthesis, screen printing, photolithography, radiation synthesis, etc [10] [16]- [22]. Here we propose the use of electrophoretic deposition method for manufacturing conductive polymer electrodes. In this work, we report for the first time, ECG recordings made with electrodes of biomimetically synthetized 3,4-ethylenedioxythiophene-Pyrrole (EDOT-PY) copolymer electrophoretically deposited onto stainless steel sheets. The SNR is close to commercial Ag/AgCl electrodes, with lower skin-impedance than Ag/AgCl electrodes.

Cyclic Voltammetry Studies
The cyclic voltametric (CV) measurements were carried out using a potentiostat

Specific Capacitance
The specific capacitance was determined by cyclovoltammetry with the configuration cell and the electrolyte mentioned above. A potential window was set at −0.7 V to 0.7 V and a sweep speed of 100 mV/s, which was calculated according to the equation where Cs is the specific capacitance, S is the integrated area of the voltametric curve (expressed as mA·V/cm 2 ), ∆U is the potential window used, and V is the sweep speed (expressed as V/s).

Scanning Electron Microscopy and Elemental Analysis
Scanning electron microscopy (SEM) images and elemental analysis (EDX) mapping were obtained using a LEO (Zeiss) 1540XB FIB scanning electron microscope with a gun emission field of 3 to 15 kV and a 5 mm working distance from the sample holder to the optical system. The EDX microanalysis was obtained using an Oxford Inca Energy 350X-MAX 50, with linear resolution of 127 Journal of Materials Science and Chemical Engineering eV in Mn Kα from 1 to 100,000 cps., The detection and quantification of elements are from atomic number 4 (Beryllium).

Transmission Electron Microscopy
The transmission electron microscopy (TEM) images were taken on a Titan Fei Thermofischer instrument with an acceleration voltage of 200 kV. The samples were prepared by immersing lacey-carbon grids in an isopropanol suspension, allowing them to dry in a vacuum oven for 2 hours at 60˚C.

Infrared Spectroscopy
Fourier transform infrared (FTIR) spectra were acquired using a Thermo Fischer Scientific FTIR Spectrophotometer in the attenuated total reflectance (ATR) mode using a diamond crystal. The sample does not require preparation. The powder is placed on the surface of the glass and the measurement is carried out.
The spectra were acquired taking an average of 32 scans with a resolution of 4 cm -1 within a range between 400 cm −1 to 4000 cm −1 .

Powder X-Ray Diffraction
The powder x-ray diffraction (PXRD) patterns were acquired using a Brucker D8 Advance Diffractometer with a Cu Kα radiation source (λ = 1.5418 Å). The powdered samples were placed in a standard sample holder. Measurements were made at intervals of 0.02˚ at a scanning speed of 10˚/min from 2θ = 2˚ to 82˚.

Electrophysiological Assays
The ECG was recorded with a MP36R Biopac unit and an RHD2000 electrode adapter plate. Electrophysiological ECG recordings were performed using an EDOT-PY 0.9 -0.1:PSS/PO 4 EPD electrode, and a Biopac electrophysiological commercial-grade Ag/AgCl electrode was used as a control. The tests were carried out with a 3-electrode configuration: the reference electrode (RE) was placed on the right ankle, the working electrode (WE) and counter electrode (CE) were placed on the left and right arm respectively, 5 cm from the wrists.
Each test used a trio of equal electrodes (flat stainless steel 1.5 × 0.5 cm electrodes). An electrophysiological gel drop was added to each electrode before starting the test. The test subject (female, average weight) remained seated and at rest while the signal was allowed to stabilize for 15 minutes, and any type of movements was avoided to not add breathing or shaking artifacts.

Signal-to-Noise Ratio Analysis
The value of the signal-to-noise ratio (SNR) was extracted from the electrophysiological ECG recordings. A MATLAB function is applied to the data to obtain The thermal noise voltage was calculated with the equation: where k is the Boltzmann's constant in J/K, T is the temperature in degrees K, B is the noise bandwidth value, and R is the impedance in Ohms.

Polymer Composition through Fourier Transform Infrared Spectrometry
Table S1 (supporting information) presents the main bands identified in the different polymers analyzed. When analyzing PEDOT-Hematin, the appearance of a band related to the asymmetric stretching C = C of the ring [23] is observed at 1539 cm −1 . Likewise, some signals associated with the symmetric stretching of the ring were also not observed, which could mean a decrease in conjugation in the chain related to defects such as branching and crosslinking, which in general terms affects the conductivity of the polymer. In general, when comparing the spectra of PEDOT or Polypyrrole (PPY) (Figure 1), most of the bands have remarkable similarity regardless of the chemical skeletons of PPY and PEDOT, which indicate that the hematin does not exert a significant effect on the chemical structure of the homopolymers of PEDOT and PPY just as a catalyst would. Figure 1 shows the Fourier transform infrared (FTIR) spectra of the EDOT-PY copolymers: 0.9 -0.1, 0.7 -0.3, 0.5 -0.5 and 0.3 -0.7 biomimetically synthesized in the presence of Hematin. In the case of the PEDOT PSS spectrum, the existing bands corroborate the chemical structure of the polymer reported in the literature. Contributions from PSS are also observed, specifically from the −SOx group, which indicates that the doping process with PSS was successful. The spectrum of PPY:PSS shows the characteristic bands observed for PPY doped with TSA as well as how groups of the polymer were observed. When analyzing the bands of the EDOT-PY:PSS copolymer (Table S1)  is only modified when the monomer molar ratio changes. Figure 1(b) presents the deposition of the biomimetically synthesized copolymer EDOT-PY 0.9 -0.1 and 0.7 -0.3 carried out on stainless steel surfaces assisted by electrophoresis. The coatings exhibit intense bands at 72:1000 and 1130 cm −1 associated with stretching and deformation of SOR and SO 2 bonds at 1250 cm −1 with CN bonds. The bands between 1330 and 1450 cm −1 are associated with stretching and deformation of the CC sp 3 and sp 2 of the heterocycle, and at 1640 cm −1 the presence of carbonyl groups. It is reasonable to conclude that the methods work for coating stainless steel because the EDOT-PY PSS/PO 4 was deposited cathodically and not anodically as would be expected. Also, the coating shows a wide and intense band centered at 1023 cm −1 , which could be associated with the co-deposition of phosphate on the surface [25], which masks or overlaps the vibrational contributions of the copolymer. Figure 1(c) shows the PEDOT, PPY and EDOT-PY diffractograms. In the case of PEDOT-PSS (2:1) the diffractions become wider because the PSS, being a semicrystalline polymer, introduces a higher level of amorphousness to the structure. Despite this, extremely intense and narrow diffraction maxima can be observed at 2θ = 4.9˚, which could indicate a significant increase in the order between adjacent segments on the XY plane [26]. In the case of PPY:PSS (2:1) the diffractions are extremely wide. However, diffraction at approximately 2θ = 30˚ does not appear when doped only with TSA, so it could be associated with chain stacking of PSS [26] [27] [28] [29]. The homopolymer and copolymer diffractograms show four widened diffractions, which indicate an increase in the Journal of Materials Science and Chemical Engineering disorder of the structure due to both, the loss of planarity introduced by the copolymer, and the presence of PSS [30]. In the case of the polyanion PSS and TSA, an interesting effect occurs in the plane associated with chain stacking (2θ = 26.2˚), since it narrows the diffraction maxima, which indicates higher order in the stacking of chains, and this could possibly increase the electrical conductivity of the material. This is likely either due to the shape and size of the polyanion nanoparticles that can promote a higher order, or that serves as nucleation and growth points for crystalline nanodomains [31]. In general, CCBS diffractograms present four wide signals centered on: 2θ = 5, 16.7, 28.6 and 41.1˚ [32] [33] [34]. These diffractions are very broad and rounded, which indicate that the PSS introduces a high degree of amorphousness to the structure. However, the 2θ = 28.6˚ diffraction maxima is sharper compared with the homopolymers probably because the structure of the copolymer generates well defined crystalline stacking microdomains. FePO 4 electrosynthesis that helps to entrap the polymer cathodically. There is a tendency for copolymer mobilization to the cathode even when the particles have a negative Z potential. One hypothesis for this is due to adsorption and ionic exchange phenomena induced by protons, acid phosphate ions and Fe 2+ cations on the surface of the polymer; nanoparticles modify the surface potential [36] and transport the polymer to the cathode surface, generating a coalescence and layer growth.  Figure 1(d)) has a typical quasi-rectangular shape of highly capacitive surfaces with low resistivity, or in other words, non-faradaic (ECDL) plus faradaic process kinetics that contributes to the net's electrochemical response of the material [37]. When compared to the voltametric curve of the uncoated Journal of Materials Science and Chemical Engineering stainless-steel electrode (blue curve, Figure 1(d)) with the voltammogram of the copolymer deposit, the minimum capacitance contributed by the metallic surface of the steel is evident. Therefore, the majority of the additional surface capacitance is attributed to the electrophoretically deposited copolymer [38]. The data above shows that the electrophoretic deposition of the EDOT-PY copolymer with PBS as an electrolyte and fixative was successful. The voltammogram obtained from a conventional Ag/AgCl electrode exhibits a capacitive behavior (ECDL), which is consistent with a high surface area and consequently a satisfactory signal-to-noise ratio (SNR). However, the voltametric evaluation shows that its potential window is very small with the PBS electrolyte (−0.045 to 0.2 V vs. Ag/AgCl), which is detrimental, since Ag/AgCl electrodes have been shown to release reactive oxygen species [39] [40] when used as a stimulation electrode in cells or microorganisms due to a narrow zone of electrochemical stability [41]. Polymer-coated electrodes should not present such problems since their operating window is wide (−0.7 to 0.7 V). In general, the specific capacitance per unit area of all the electrodes analyzed in Table 1 and Figure 2    concerning PEDOT, which indicates an increase in the number of polymer particles deposited on the surface and a decrease in phosphorous-sodium complexes associated with carry-over and coalescence on the electrode. The reason fundamentally may be due to the size of the copolymer particle that is smaller regarding PEDOT (~30 nm vs. 200 nm, see TEM characterization). In general, it is homogeneous, although micrometric grooves are observed which generate laminar segments along with the electrode, which could generate long-term delamination. In the cycles evaluated in this study, no loss of electrode integrity or variations in resistance and signal-to-noise ratio were observed (see Conductivity and SNR below), which indicate that it is only a morphological effect of the macro-relief and obeys the mobility kinetics. shown that the copolymer with a small molar fraction of pyrrole drastically modifies its colloidal properties. The effect is corroborated by the decrease in particle size and the racemic morphology observed of multiple nanoparticles between 10 and 20 nm that form micrometric clusters. In Figure 3(b), we can see morphological characteristics that could be related to the nanometric detail of Figure 3(e). The electrophoretic deposit drives the massive coalescence and agglomeration of cathodically transported nanoparticles which eventually becomes a highly porous coating.

Electrocardiogram Assays
Electrocardiogram recordings were performed using the methodology described   distinguished that [44]: the P wave corresponds to atrial depolarization, the QRS complex is associated with massive ventricular depolarization. The peak constitutes the highest signal intensity due to the massive contribution of ventricular fibers that reflect millions of action potentials of individual cardiomyocytes. After this event, the T wave is distinguished, which corresponds to ventricular repolarization. The frequencies between the P and Q waves are within 180 -190 ms, while the S-T frequency is for less than 400 ms, which indicates that there are no AV blocks. In general, the quality of the electrocardiogram is sufficiently satisfactory for diagnostic purposes [7]. The upper red plot of Figure 4 corresponds to the electrocardiogram recorded simultaneously with Ag/AgCl Biopac contact electrodes. The recorded signals are less wide, the amplitude of the QRS complex wave is lower, and the sinusoidal shape of the P and T waves was observed. This is verified when performing the electronic analysis of the signals; both electrodes have a noise RMS close to 6 µV ( Figure 6(c)). However, the SNR of the EDOT-PY electrode is equal to 36 dB while the Ag/AgCl electrode is close to 24 dB (Figure 6(b)).
This implies a better signal-to-noise ratio for the conductive copolymer electrode, which is strongly associated with surface conductivity and electrode capacitance. The copolymer coated electrode exhibits a surface conductivity of 0.08 S*sq and capacitance of 2.02 F/cm 2 against a surface conductivity of 0.25 S*sq and a specific capacitance of 0.5 F/cm 2 for the Ag/AgCl electrode. In a very gen-Journal of Materials Science and Chemical Engineering eral sense, it can be argued that CCBS copolymer electrodes have lower skin contact impedance than the commercial electrodes mostly used in commercial assays ( Figure 6(a)). The results reflect differences in surface conductivity and capacitance per unit area. It has a significant effect on electrophysiological recordings and paves the way for future tests for stimulation and recording of electrical activity in neurons. Figure 5 shows the ECG recordings obtained for the II, III and AVR leads.
The same protocol described above was followed in this case. The only difference was the position of the electrodes on the test subject. The ECG segment shows typical signals for all recordings: P wave, QRS complex, and T90 wave.
The recording is satisfactory for all leads. Not much thermal noise is observed ( Figure 6(c)), and neither is the signal amplitude; for all of them is similar concerning the first lead signal amplitude. The electroactive area is the same for all electrodes and therefore the density of signals collected per unit area [45]. Any  observe that the first, second, and third ECG leads exhibit similar skin impedance, but the AVR lead is lower than the other recordings. This implies that the position of the electrodes is highly dependent on the skin impedances obtained by CCBS electrodes as well as by Ag/AgCl electrodes. However, the thermal noise voltage remains low for all ECG recordings.
We believe that both kinds of electrodes are sufficiently accurate for low noise electrocardiogram recording which would imply that the CCBS electrodes are as useful and effective as commercial Ag/AgCl ECG as discussed above. Einthoven's lead I and recording one ECG for two hours with normal subject test mobility throughout the assay. After that, the electrode was washed with water and ethanol several times and dried, then the same protocol was repeated two more times. We found that the SNR decreased in a monotonical way after each cycle. The electrode held 80% of the initial SNR value after being reused three times, however the SNR value was still high and acceptable for ECG recording with low noise. This implies that the CCBS electrodes probably work for long-term recordings or wearable applications. Journal of Materials Science and Chemical Engineering

Conclusion
The Ag/AgCl electrodes exhibited a low specific capacitance per unit area (0.5 F/cm 2 ) as well as a narrow working potential window (−0.045 to 0.2 V vs. Ag/AgCl) compared with the CCBS electrodes manufactured in this work by electrophoretic deposition, which display high capacitances and a wide working potential window (−0.7 to 0.7 V). It is possible to conclude that the EDOT:PY copolymer electrodes fabricated here possess high stability, which is one of the most desirable properties in long-term implants. Finally, electrophysiological tests showed high sensitivity to bioelectric signals when EDOT-PY 0.9 -0.1/PO 4 copolymer electrodes were used, which exhibit low thermal noise (6 µV) and a high SNR (36 -24 dB). This allowed us to obtain defined electrocardiographic traces. The conductive copolymer material synthesized here is ideal for the medical diagnosis of cardiac disorders or musculoskeletal diseases. This work constitutes the first report of conductive copolymer electrodes manufactured by electrophoretic deposition successfully applied to electrophysiology recordings.