nanomaterials Article

Graphene-Enabled Electrodes for Electrocardiogram Monitoring Numan Celik 1, *, Nadarajah Manivannan 1 , Andrew Strudwick 2 and Wamadeva Balachandran 1 1

2

*

Department of Electronics and Computer Engineering, Brunel University London, Kingston Lane, Uxbridge UB8 3PH, UK; [email protected] (N.M.); [email protected] (W.B.) 2-Dtech, Core Technology Facility, Incubator Building, 46 Grafton Street, Manchester M13 9NT, UK; [email protected] Correspondence: [email protected]; Tel.: +44-189-526-8459

Academic Editors: Chen-Zhong Li and Ling-Jie Meng Received: 30 June 2016; Accepted: 12 August 2016; Published: 23 August 2016

Abstract: The unique parameters of graphene (GN)—notably its considerable electron mobility, high surface area, and electrical conductivity—are bringing extensive attention into the wearable technologies. This work presents a novel graphene-based electrode for acquisition of electrocardiogram (ECG). The proposed electrode was fabricated by coating GN on top of a metallic layer of a Ag/AgCl electrode using a chemical vapour deposition (CVD) technique. To investigate the performance of the fabricated GN-based electrode, two types of electrodes were fabricated with different sizes to conduct the signal qualities and the skin-electrode contact impedance measurements. Performances of the GN-enabled electrodes were compared to the conventional Ag/AgCl electrodes in terms of ECG signal quality, skin–electrode contact impedance, signal-to-noise ratio (SNR), and response time. Experimental results showed the proposed GN-based electrodes produced better ECG signals, higher SNR (improved by 8%), and lower contact impedance (improved by 78%) values than conventional ECG electrodes. Keywords: graphene; electrocardiogram (ECG); long-term monitoring; dry electrodes; contact impedance; Raman spectroscopy

1. Introduction Cardiovascular diseases are the leading cause of death worldwide, killing more than 15 million people every year [1]. Continuous and real-time monitoring of the heart activities with well-established medical test electrocardiography (ECG) plays a very important role in early diagnosis and treatment of cardiovascular diseases. The ECG waveform that reflects the electrical activity of the heart is widely used for early detection and management of heart problems. Traditionally, the most commonly used bioelectrodes are of the gel-type disposable Ag/AgCl electrodes [2] and these are simple, reliable, and cost effective during cardiac monitoring. These standard wet-type electrodes typically have three main parts: Ag/AgCl layer for sensing; conductive gel to maintain good electrical contact with the skin; and a connector part for conducting the electrical signal through third-party devices for monitoring purposes. On the other hand, it is not suggested to use these wet electrodes in long-term cardiac monitoring due to skin irritation and allergic reactions. Studies have showed that using adhesive gel on wet Ag/AgCl electrodes can trigger dermal irritation and cause potential signal degradation in long-term measurements [3]. Therefore, the demand for comfortable and easy-to-use electrodes has resulted in fabrication of dry or gel-less ECG electrodes that eliminate the need for gel and even skin preparation.

Nanomaterials 2016, 6, 156; doi:10.3390/nano6090156

www.mdpi.com/journal/nanomaterials

Nanomaterials 2016, 6, 156

2 of 16

Several attempts have been made in fabrication of dry-contact ECG electrodes made of metallic and silicon substrates that penetrate the skin surface for electrophysiological measurements [4–6]. Although this method is clinically invasive and gel-less contact with the skin, it is difficult to satisfy electrical stability between skin and the electrode. Therefore, the applications with dry-contact ECG electrodes need to be investigated further for long-term measurements. Another promising approach is based on capacitive coupling using non-contact type dry ECG electrodes [7,8]. These types of electrodes are classified as a non-contact because the electrodes are isolated with an insulating layer from the skin with hair, clothing, air, or other synthetic or textile fabrics. In non-contact electrodes, the risk of skin irritation and preparation is eliminated; the deployment of the electrodes to the skin is easy and safe, and integration with other conductive fabrics or textile is possible. However, they have issues with contact impedance and motion artifacts throughout capacitive-coupling ECG signal from the body [9]. To overcome these limitations of dry electrodes, researchers have investigated the use of nanomaterial-based electrodes in long-term ECG sensing due to their properties of high electrical conductivity and flexibility for better wearable electronics. Myers et al. [10] developed a silver nanowire-based (AgNW) dry electrode for ECG monitoring where the NWs were integrated below the surface of the polydimethylsiloxane (PDMS) layer on the dry electrode. The electrode was used on a wristband with a conductivity of ~5 × 105 S/m at 50% tensile strain. The skin–electrode contact impedance was decreased by applying a mild pressure onto the AgNW electrode, and the results demonstrated that the AgNW/PDMS dry electrode outperformed the pregelled Ag/AgCl electrodes with fewer motion artifacts when the subject was swinging their arms and jogging. Due to interesting mechanical and electrical properties of carbon-based nanomaterials, carbon nanotubes (CNTs) are selected to disperse into the flexible polymer substrates to make a better conductive ECG electrode. Lee and his colleagues [11] fabricated a dry ECG electrode by dispersion of CNT into the PDMS layer to investigate the effects of CNT concentration and robustness to motion and sweat. Tests showed that there was no sign of signal degradation after continuously wearing the CNT/PDMS electrodes after exercise (in the presence of sweat), and for 7 days. In addition, the results revealed that there was no skin irritation after wearing the electrodes on the forearms for 7 days. However, the main point for developing flexible electrodes lies in flexible substrate and conductive material. Conventional methods such as gel, contact, and non-contact dry-type of ECG electrodes tend to have low charge carrier density, thus causing low conductivity and lower signal quality. Although nanomaterials-based electrodes provide a better signal-to-noise ratio (SNR) with flexibility, and eliminate skin irritation problems, there are still shortcomings to be resolved in long-term ECG monitoring for better accuracy. Previous studies indicated that metals deposited on PDMS exhibited wrinkles and cracks frequently [12,13]. Additionally, it is difficult to achieve a proper mixing process due to the stickiness of polymer. Besides, an experiment showed that the adhesion between metallic layers and PDMS substrate is poor [14]. More importantly, the conductivity of the conductive polymer is not as good as a metal conductor (80 S/m versus 5 × 105 S/m) [5]. For further development of conductive and flexible electrodes into wearables, it is important to use highly conductive (~107 S/m) and thin metals in such a reliable, stable, and long-term ECG sensing application. Graphene (GN), a single-layer two-dimensional structure nanomaterial, exhibits exceptional physical, electrical, and chemical properties that lead to many applications from electronics to biomedicine. The unique parameters of GN—which has notably the highest electron mobility (~200,000 cm2 ·V−1 ·s−1 ), the highest thermal conductivity (5300 W·m−1 ·K−1 ), the highest surface area to volume ratio (2630 m2 /g), the fastest moving electrons (~106 m/s), and is the best conductor of electricity (resistivity of 10−6 Ω·cm) in any material—are bringing heightened attention into biomedical applications [15]. In this study, we fabricated a dry electrode by growing GN within copper (Cu) substrate on top of a silver (Ag) layer of a commercial ECG electrode using a chemical vapour deposition (CVD) method. Based on this technique, we produced a GN-based ECG electrode for the first time on the development of ECG electrodes with GN substrates. We measured contact impedance according to frequency changes and compared the results with those of conventional Ag/AgCl

Nanomaterials 2016, 6, 156

3 of 16

electrodes. Furthermore, we measured ECG signals and quantitatively compared the performance of the new GN-based ECG electrodes against wet electrodes. To determine the feasibility of long-term Nanomaterials 2016, 6, 156 3 of 16 monitoring, we studied the influence of GN-based electrodes on 10 people. We tested these electrodes for one person inmonitoring, each week we andstudied demonstrated theirofrobustness in ECG monitoring. of long-term the influence GN-based electrodes on 10 people. We tested these electrodes for one person in each week and demonstrated their robustness in ECG monitoring.

2. Materials and Methods 2. Materials and Methods

2.1. Fabrication Process 2.1. Fabrication Process

A number of different algorithms for synthesizing GN have been reported since first obtained number of different algorithms formechanical synthesizingexfoliation, GN have been reported since first obtained in 2004 byANovoselov and Geim, including liquid-based exfoliation, epitaxial in 2004 by Novoselov and Geim, including mechanical exfoliation, liquid-based exfoliation, epitaxial growth on silicon carbide, and chemical vapour deposition (CVD) growth GN on metal substrates growth on silicon carbide, and chemical vapour deposition (CVD) growth GN on metal substrates such as silicon, nickel, and copper [15]. To fabricate a GN-enabled conductive electrode, we have used such as silicon, nickel, and copper [15]. To fabricate a GN-enabled conductive electrode, we have the CVD approach (Figure 1) for coating GN on top of a metallic substrate of the commercial ECG used the CVD approach (Figure 1) for coating GN on top of a metallic substrate of the commercial electrode. This has advantages on transition into different metals metals such assuch Ni, as Cu,Ni, and substrates, ECG electrode. This has advantages on transition into different Cu,Si and Si and CVD-grown GN is inexpensive and a feasible method for single-layer GN synthesis compared substrates, and CVD-grown GN is inexpensive and a feasible method for single-layer GN synthesis to othercompared production techniques. to other production techniques.

Figure 1. Schematic diagram showingcoating coating process process for graphene-coated electrode. Figure 1. Schematic diagram showing forsynthesizing synthesizing graphene-coated electrode.

We coated GN as a layer onto the target electrodes after GN growth on a metallic substrate and

We coated GN aswhich a layer onto the target GN growth a metallic transfer processes, were carried out byelectrodes a specialistafter graphene companyon[16]. The GN substrate was and transfer which (H were carried out by4) agases specialist graphene company [16]. The GN producedprocesses, utilising hydrogen 2) and methane (CH based on the chemical vapour deposition (CVD) growth procedure, similar (H to those used elsewhere in 4the literature on copper substrates was produced utilising hydrogen methane (CH ) gases based on the(Cu) chemical vapour 2 ) and [17–19](CVD) to yieldgrowth monolayer graphenesimilar films. These films were subsequently transferred to on a Ag layer (Cu) deposition procedure, to those used elsewhere in the literature copper of the [17–19] target electrodes a wet chemical approach etch away Cu substrate, involving a spin- to a substrates to yield via monolayer graphene films.toThese filmsthe were subsequently transferred coated Polymethyl methacrylate (PMMA) support layer which is later removed from the GN–Ag Ag layer of the target electrodes via a wet chemical approach to etch away the Cu substrate, involving a substrate form with solvents, akin to work carried out elsewhere [20,21]. The details of growth and spin-coated Polymethyl methacrylate (PMMA) support layer which is later removed from the GN–Ag transfer processes can be seen in Figure 1, which shows the fabricated electrode within a very thin substrate form with solvents, akin to work carried out elsewhere [20,21]. The details of growth and GN layer—which was about 3.7 Å (0.37 nm)—for ECG sensing applications. Electrical properties of transfer processes be seenwere in Figure 1, which shows the fabricated within a very thin the GN-coatedcan electrodes measured at room temperature with aelectrode two-probe method using a GN layer—which was about Å (0.37 nm)—for sensing (σ) applications. Electrical properties digital multimeter (0.13.7 Ω–2.8 Ω). The electricalECG conductivity of the electrode before and after GNof the GN-coated were at room temperature with a two-probe method using a digital coatingelectrodes was measured as measured 2.48 × 102 S/m and 6.94 × 104 S/m, respectively. multimeter (0.1 Ω–2.8 Ω). The electrical conductivity (σ) of the electrode before and after GN coating 2 S/m 2.2. Raman as Spectroscopy SEMand Images was measured 2.48 × 10and 6.94 × 104 S/m, respectively. GN coating process can be repeatable on a variety of common metallic crystals such as copper

2.2. Raman Spectroscopy and dioxide SEM Images (Cu), nickel (Ni), silicon (SiO2), platinum (Pt), gold (Au), and silver (Ag). Cu is ideal metallic crystal for GN growth of its low and the carbon dissolution at high (Cu), GN coating process canbecause be repeatable on acarbon varietysolubility, of common metallic crystals such as copper temperature into the bulk helped in self-termination of monolayer growth [22]. Furthermore, Cu is nickel (Ni), silicon dioxide (SiO2 ), platinum (Pt), gold (Au), and silver (Ag). Cu is ideal metallic crystal selected here for fabrication of the ECG electrodes due to the high deposition of GN while transferring for GN growth because of its low carbon solubility, and the carbon dissolution at high temperature GN onto the Ag substrates of the electrodes. Raman spectroscopy and scanning electron microscopy

Nanomaterials 2016, 6, 156

4 of 16

into the bulk helped in self-termination of monolayer growth [22]. Furthermore, Cu is selected here for fabrication of the ECG electrodes due to the high deposition of GN while transferring GN onto the Ag substrates of the electrodes. Raman spectroscopy and scanning electron microscopy (SEM) analyses were done in order confirm the presence of the coating of GN on the fabricated electrode after Nanomaterials 2016,to 6, 156 4 of 16 completing the transferring process. SEM analysis was done by the Zeiss Supra field emission electron (SEM) analyses were done in order to confirm the presence of the coating of GN on the fabricated gun (FEG)-SEM and Raman spectroscopy analysis was done using a confocal Raman spectrometer electrode after completing the transferring process. SEM analysis was done by the Zeiss Supra field that was available at 2-D Tech [16]. emission electron gun (FEG)-SEM and Raman spectroscopy analysis was done using a confocal As Raman can bespectrometer seen fromthat Figure 2, GN at has was available 2-Dexpanded Tech [16]. through almost the entire cross-section of the electrode, which wasFigure designed increase thethrough contact areathe between the electrode and skin. As and can be seen from 2, GNtohas expanded almost entire cross-section of the electrode, and which was were designed to increase the contact area between theafter electrode and skin.process. These SEM These porous structure layers observed under the SEM before and GN coating porous structureelectrode layers were observed the SEMexperimental before and after GN coating process. SEMpatterns images of GN-coated were taken under after several studies. Because of this, images of GN-coated electrode were taken after several experimental studies. Because of this, of residual points were detected just above the GN layer on the surface of the electrode (as white colour patterns of residual points were detected just above the GN layer on the surface of the electrode (as in the Figure 2). These images further show that GN coating has smoothed the surface and hence can white colour in the Figure 2). These images further show that GN coating has smoothed the surface improveand thehence skin-to-electrode and detect better signals. can improve thecontact skin-to-electrode contact andECG detect better ECG signals.

Figure 2. Scanning electron micrographs (SEMs) showing: (a) and (b) SEM of dry electrode before

Figure 2. Scanning electron micrographs (SEMs) showing: (a,b) SEM of dry electrode before coating; coating; (c) and (d) graphene-coated dry electrode. (c,d) graphene-coated dry electrode. Optical detection relying on light scattering is especially attractive because it is fast, sensitive, and nondestructive. In particular, spectroscopy analysis is the most powerful Optical detection relying on lightRaman scattering is especially attractive because it istechnique fast, sensitive, available for the characterization of carbon nanostructures such as nanotubes and graphene [23]. The and nondestructive. In particular, Raman spectroscopy analysis is the most powerful technique Raman spectrum of carbon-based materials shows two main peaks, the G- and D-peaks, which lie at available for the characterization of carbon nanostructures such as nanotubes and graphene [23]. around 1580 cm−1 and 1360 cm−1, respectively (Figure 3b). The G-peak corresponds to the E2g phonon The Raman of carbon-based materials shows two main peaks, G- and D-peaks, 2 atoms at thespectrum Brillouin zone center. The D-peak is due to the breathing modes of spthe and requires a which − 1 − 1 lie at around 1580 cm and 1360 cm the , respectively (Figure The G-peak corresponds defect for its activation [24]. However, most prominent feature3b). in graphene is the second order of to the the D-peak: 2D-peak.zone This lies at around cm−1 (Figure is always seen, evenof when E2g phonon at thethe Brillouin center. The2700 D-peak is due3b) to and the itbreathing modes sp2 atoms no D-peak is present, no defects required the for the activation of second-order phonons. is the and requires a defect for its since activation [24].are However, most prominent feature in graphene A supplementary material file has been added to this work as it shows the Raman analysis of bare − 1 second order of the D-peak: the 2D-peak. This lies at around 2700 cm (Figure 3b) and it is always seen, even when no D-peak is present, since no defects are required for the activation of second-order

Nanomaterials 2016, 6, 156

5 of 16

Nanomaterials 2016, 6, 156

5 of 16

phonons. A supplementary material file has been added to this work as it shows the Raman analysis of bare Ag/AgCl and Graphene-coated respectively Figure S1, andS2Figure S2 Ag/AgCl electrodeelectrode and Graphene-coated electrodeelectrode respectively in FigureinS1, and Figure in-detail in-detail for detecting in 10 measurements. Its shape and position distinguishes single-layer for detecting 2D band2D inband 10 measurements. Its shape and position distinguishes single-layer from from multilayer samples. Regarding our tests, the Raman spectra of GN and GN-coated sample was multilayer samples. Regarding our tests, the Raman spectra of GN and GN-coated was measured by taking 10 point measurements across the whole sample surface at 488 nm excitation. In measured by taking 10 point measurements across the whole sample surface at 488 nm excitation.an In attempt to get clearer datadata andand rulerule out any fluorescence signals fromfrom the underlying substrate, 10 data an attempt to get clearer out any fluorescence signals the underlying substrate, 10 points were subsequently obtained using using a 514 nm laser. 3 shows the Raman spectrum of GN data points were subsequently obtained a 514 nm Figure laser. Figure 3 shows the Raman spectrum grown Cu (Figure and for electrode and bare Ag/AgCl 3b) of GN on grown on Cu3a) (Figure 3a)both andGN-coated for both GN-coated electrode and bareelectrode Ag/AgCl(Figure electrode obtained from the 488from nm the laser. can be seen from images, three important (G, D, 2D) (Figure 3b) obtained 488Asnm laser. As can bethe seen from the images, three bands important bands were Raman spectrum separately selected to be analysed. (G, D,observed 2D) wereinobserved in Ramanand spectrum and 2D-peaks separatelywere 2D-peaks were selected to be analysed.

Figure 3. 3. Raman Raman analysis analysis of of graphene graphene (GN) (GN) and andbare bareAg/AgCl Ag/AgCl electrode. electrode. (a) (a) Raman Raman spectrum spectrum of of Figure graphene grown on Cu after chemical vapour deposition (CVD); and (b) Raman spectrum for the graphene grown on Cu after chemical vapour deposition (CVD); and (b) Raman spectrum for the GN-coated electrode electrode and and conventional conventional Ag/AgCl Ag/AgCl bare GN-coated bare electrode. electrode.

According to Figure 3a, Raman spectra was measured for the single-layer graphene (GN) area According to Figure 3a, Raman spectra was measured for the single-layer graphene (GN) area for for the sample which was grown on Cu after CVD. A shift in the 2D-peak position of single-layer GN the sample which was grown on Cu after CVD. A−1shift in the 2D-peak position of single-layer GN has has been observed in the range of 2711–2723−1cm . The value of full width at half maximum varies in been observed in the range of 2711–2723 cm . The value of full width at half maximum varies in the the range of 25–31 cm−1 which represents the single layer of GN [25]. The ratio of 2D/G also indicates range of 25–31 cm−1 which represents the single layer of GN [25]. The ratio of 2D/G also indicates that the sample was grown by single-layer GN, and this ratio for this sample has been varied in the that the sample was grown by single-layer GN, and this ratio for this sample has been varied in the range of 5–7. Figure 3b shows the comparison of Raman spectra for the GN-coated electrode and range of 5–7. Figure 3b shows the comparison of Raman spectra for the GN-coated electrode and conventional bare Ag/AgCl electrode. As can be seen from this figure, the presence of the 2D-peak in conventional bare Ag/AgCl electrode. As can be seen from this figure, the presence of the 2D-peak in the GN-coated electrode (in red square) differentiates the presence of GN from the bare electrode. the GN-coated electrode (in red square) differentiates the presence of GN from the bare electrode. 2.3. Measurements of ECG Acquisition System ECG signals were separately measured from proposed GN-based electrodes and conventional pre-gelled Ag/AgCl electrodes (Ambu) via ECG acquisition system (e-Health and Arduino) using

Nanomaterials 2016, 6, 156

6 of 16

2.3. Measurements of ECG Acquisition System ECG signals were separately measured from proposed GN-based electrodes and conventional Nanomaterials 2016, 6, 156electrodes (Ambu) via ECG acquisition system (e-Health and Arduino) 6 of 16 pre-gelled Ag/AgCl using 3-leads to allow measurements from multiple biopotential electrodes. Figure 4 shows both types of 3-leads to allow measurements from multiple biopotential electrodes. Figure 4 shows both types of electrodes (GN-based and conventional pre-gelled ECG electrodes) used in the experiments. In order electrodes (GN-based and conventional pre-gelled ECG electrodes) used in the experiments. In order to obtain the electrocardiogram, ECG lead-1 configuration was conducted throughout experimental to obtain the electrocardiogram, ECG lead-1 configuration was conducted throughout experimental testing of ECG electrodes on a 29 year old male; two active electrodes were attached to the left and right testing of ECG electrodes on a 29 year old male; two active electrodes were attached to the left and chest and the other driven-right-leg (DRL) electrode was attached onto the left waist for referencing. right chest and the other driven-right-leg (DRL) electrode was attached onto the left waist for ECG signals were filtered between 0.5 and 100 Hz and an additional notch filter was applied at 50 Hz referencing. ECG signals were filtered between 0.5 and 100 Hz and an additional notch filter was cutoff frequency during sampling and amplifying of ECG signals at the signal acquisition system, applied at 50 Hz cutoff frequency during sampling and amplifying of ECG signals at the signal which is shown in Figure 4. We have coated GN on two types of electrodes which were (1) slightly acquisition system, which is shown in Figure 4. We have coated GN on two types of electrodes which bigger typebigger of electrode smaller size electrode. 4a,b shows wereAmbu (1) slightly Ambu and type(2) of electrode andCovidien (2) smallertype size of Covidien typeFigure of electrode. Figurethe Ambu type of electrode just before and after GN coating, respectively. This type of electrode 4a,b shows the Ambu type of electrode just before and after GN coating, respectively. This typehas of a diameter of 38 mm and a thickness of 1 mm. Likewise, the Covidien type of electrode is shown electrode has a diameter of 38 mm and a thickness of 1 mm. Likewise, the Covidien type of electrode in Figure 4c,d in just before and after GNand coating, respectively, and this type ofthis electrode diameter is shown Figure 4c,d just before after GN coating, respectively, and type ofhas electrode has of 24 diameter mm and aofthickness of 1 mm. ECG signals were measured after amplifying, filtering, and sampling 24 mm and a thickness of 1 mm. ECG signals were measured after amplifying, filtering, processes using aprocesses data acquisition platform whichplatform is also shown Figure 4e. in Figure 4e. and sampling using a data acquisition which in is also shown

Figure Experimentallyused usedelectrodes electrodesand and data data acquisition electrode (Ambu) Figure 4. 4. Experimentally acquisitionsystem. system.Bigger Biggersize sizeofof electrode (Ambu) before GN coating in (a), and after GN coating in (b); smaller size of electrode (Covidien) before GN before GN coating in (a), and after GN coating in (b); smaller size of electrode (Covidien) before GN coating in (c), and after GN coating in (d); data acquisition unit in (e). coating in (c), and after GN coating in (d); data acquisition unit in (e).

The mechanism of graphene-based electrode is shown in Figure 5a as a separate graphene layer The graphene-based electrode shown inlayer Figure 5a as a separate graphene layer and Agmechanism layer beforeofand after the coating process.isGraphene is designed for the purpose of the and Ag layer before and after the coating process. Graphene layer is designed for the purpose of skin–electrode contact area to get the utmost ECG signal even from behind the ear location. Graphenethe skin–electrode contact areatype to get ECG signal behindonthe location. Graphene was coated onto the dry of the Ag utmost layer electrode. The even use offrom graphene dryear ECG electrodes has was coatedadvantages: onto the dry type of Ag electrode. Thehaving use of gel); graphene on dry ECG electrodes has several technically lesslayer complex (without physically attached to non-hair several advantages: technically complex (without having gel); attached to non-hair regions, hence more suitable forless long term use; and user friendly, as physically there is no need to remove the regions, hence more for long term use; and user friendly, there no need toworks remove top garment. On thesuitable other hand, the conventional Ag/AgCl electrodeasfor ECGismonitoring withthe works as athe conductive layer Ag/AgCl between the skin andfor electrode to lower skin– topelectrolyte garment. gel Onwhich the other hand, conventional electrode ECG monitoring works electrode contact andas stablise receptionlayer of the signal. the However, the electrode electrolytetogellower is with electrolyte gel impedance which works a conductive between skin and undesirable for long-term ECG monitoring due to its causing skin-related problems such as skin–electrode contact impedance and stablise reception of the signal. However, the electrolyte gel is dermatitis.for Moreover, electrolyte gel dries time, which could result insuch degradation of undesirable long-termthe ECG monitoring due toout itswith causing skin-related problems as dermatitis. signal quality and so prevents body for thecould patient at the time. Toof overcome the Moreover, the electrolyte gel dries out movements with time, which result insame degradation signal quality drawbacks, the graphene-coated dry electrodes are suggested for long-term ECG monitoring with and so prevents body movements for the patient at the same time. To overcome the drawbacks, wirelessly transmission ECG data in this concept (see Figure 5b).

Nanomaterials 2016, 6, 156

7 of 16

the graphene-coated dry electrodes are suggested for long-term ECG monitoring with wirelessly transmission ECG data in this concept (see Figure 5b). Nanomaterials 2016, 6, 156 7 of 16

Figure 5. The mechanism of graphene-coated electrode; (b) Figure 5. The Thegraphene-electrode graphene-electrodemechanism: mechanism:(a)(a) The mechanism of graphene-coated electrode; the schematic illustration of the developed system while ear-lead electrocardiogram (ECG) was tested (b) the schematic illustration of the developed system while ear-lead electrocardiogram (ECG) was on the on subject, including a wireless ECG data system. tested the subject, including a wireless ECGacquisition data acquisition system.

Furthermore, the skin–electrode contact impedance of the GN-coated electrode was measured Furthermore, the skin–electrode contact impedance of the GN-coated electrode was measured and compared with that of the conventional Ag/AgCl pre-gelled electrode by an impedance analyser and compared with that of the conventional Ag/AgCl pre-gelled electrode by an impedance analyser (Wayne Kerr 6500B (Wayne Kerr Electronics, West Sussex, UK)). It is known that a better ECG signal (Wayne Kerr 6500B (Wayne Kerr Electronics, West Sussex, UK)). It is known that a better ECG signal would give lower skin-to-electrode contact impedance because of high quality signal acquisition would give lower skin-to-electrode contact impedance because of high quality signal acquisition comprised of minimal noise. Results will be specified later, demonstrating the attractive performance comprised of minimal noise. Results will be specified later, demonstrating the attractive performance of the GN-enabled electrode for measurements of the ECG and skin–electrode contact impedance. of the GN-enabled electrode for measurements of the ECG and skin–electrode contact impedance. 3. Experimental Results and Discussion 3. Experimental Results and Discussion 3.1. Simulation Electroplating Graphene Graphene 3.1. Simulation of of Electroplating Since the the Ag Ag layer graphene Since layer has has already already been been placed placed on on the the electrode, electrode, we we wanted wanted to to electroplate electroplate graphene (GN) on top of electrode to increase the surface electrical properties and to see the differences using (GN) on top of electrode to increase the surface electrical properties and to see the differences using finite element elementsoftware softwareComsol ComsolMultiphysics. Multiphysics. Figure 6 draws design of skin model each finite Figure 6 draws the the design of skin model with with each layer layer of the skin. Considering the fact that physiological systems are very complex in nature, here we of the skin. Considering the fact that physiological systems are very complex in nature, here we only only considered the following parameters in the construction of our skin model: (1) Stratum considered the following parameters in the construction of our skin model: (1) Stratum Corneum; Corneum; (2) Epidermis; and ECG (4) Muscle. ECG canby beplacing required by placing on a (2) Epidermis; (3) Dermis; (3) andDermis; (4) Muscle. can be required electrodes on electrodes a person’s skin, person’s skin, and coupling between electrode and skin could be described as a resistor and a and coupling between electrode and skin could be described as a resistor and a capacitor in parallel [26]. capacitor in parallel [26]. The skin–electrode contact impedance is dependent on electrical The skin–electrode contact impedance is dependent on electrical conductivity or resistivity of each conductivity orassigned resistivity each layer. Here, wevalues assigned the electrical conductivity values of layer. Here, we theofelectrical conductivity of Stratum Corneum, Epidermis, Dermis, Stratum Corneum, Epidermis, Dermis, and Artery (Blood) as 0.0005 S/m, 0.95 S/m, 0.2 S/m, and 0.8 and Artery (Blood) as 0.0005 S/m, 0.95 S/m, 0.2 S/m, and 0.8 S/m (see in Table 1), respectively, S/m (see in Table 1),mode; respectively, in simulation of thefrom mode; these values werefrom obtained fromthere [27]. in simulation of the these values were obtained [27]. As can be seen the figure, As can be seen from the figure, there is a bottom layer of the electrode called referenced graphene. is a bottom layer of the electrode called referenced graphene. We filled in that layer with Cu first, and We filled in that layerto with Cu the first, and then graphene in ordersimulations to compare theperformed electrical then graphene in order compare electrical characteristics. Again, were characteristics. simulations were Ag/AgCl performed forgraphene-coated two types of Ag/AgCl. electrodes: The conventional for two types ofAgain, electrodes: conventional and values of Ag/AgCl and graphene-coated Ag/AgCl. The values of electrical conductivity were applied as 6 × 107 7 9 electrical conductivity were applied as 6 × 10 S/m and 1 × 10 S/m for Cu and GN, respectively. S/m and 1 × 109 S/m for Cu and GN, respectively.

Nanomaterials 2016, 6, 156 Nanomaterials 2016, 6, 156 Nanomaterials 2016, 6, 156

8 of 16 8 16 of 16 8 of

Figure 6. Initial skin model with two electrodes constructed in Comsol.

Figure with two two electrodes electrodesconstructed constructedininComsol. Comsol. Figure6.6.Initial Initialskin skin model model with Table 1. Electrical conductivity and relative permittivity values of each layer in Comsol modelling.

Table 1. 1.Electrical valuesofofeach eachlayer layerininComsol Comsol modelling. Table Electricalconductivity conductivityand and relative relative permittivity permittivity values modelling. Layers of Modelling Electrical Conductivity (S/m) Relative Permittivity Layers of Modelling Electrical Conductivity (S/m) Relative Relative Permittivity Graphene (GN) 1.18 × 109 (S/m) 11.5 Layers of Modelling Electrical Conductivity Permittivity 9 6 Graphene (GN) 1.18 11.5 Ag-Silver 61.6 ×× 10 10 3.4 Graphene (GN) 11.5 1.18 × 109 6 Ag-Silver 61.6 × 10 6 Stratum Corneum 0.0005 Ag-Silver 3.4 13.4 61.6 × 10 Epidermis 0.95 Stratum Corneum 0.0005 Stratum Corneum 0.0005 1 11 Epidermis 0.95 1 11 Dermis 0.2 Epidermis 0.95 Dermis 0.2 1 1 Artery 0.8 Dermis 0.2 1 Artery

Artery

0.8

0.8

1

1

After GN electroplating, the concentration distribution of the electromagnetic field was depicted inAfter Figure 7.electroplating, The electrical field effect showed the electrical distribution of an artery,field which is depicted drawn After GN the distribution ofthe theelectromagnetic electromagnetic field was depicted GN electroplating, theconcentration concentration distribution of was in Figure 7, and the capture effect of the GN-filled electrode on top of the skin layer within current in in Figure 7. 7.The the electrical electricaldistribution distributionofofanan artery, which is drawn Figure Theelectrical electricalfield fieldeffect effect showed showed the artery, which is drawn density lines.the It was also analysed in Comsol that theelectrode surface charge density ofskin the GN-coated bottom in in Figure 7,7,and ontop top the layer within current Figure and thecapture captureeffect effectof of the the GN-filled GN-filled electrode on ofofthe skin layer within current layer was higher than the Ag-filled electrode, thus capturing more electrical potential from the density lines. It was also analysed in Comsol that the surface charge density of the GN-coated bottom density lines. It was also analysed in that the surface charge density of the GN-coated bottom surface (see Figure 7a,b). The electromagnetic field and distribution was also depicted by measuring layer was higher than the the Ag-filled electrode, thusthus capturing moremore electrical potential fromfrom the surface layer was higher than Ag-filled electrode, capturing electrical potential the surface electric displacement field. It was visible that the GN deposition at the bottom of the surface (see Figure 7a,b). The electromagnetic field and distribution was also depicted by measuring (see Figure 7a,b). The electromagnetic field and distribution was also depicted by measuring surface electrodes was capturing electrical potential due to higher current density.

surface electric displacement field. It was the GN at deposition at of thethe bottom of the electric displacement field. It was visible thatvisible the GNthat deposition the bottom electrodes was electrodes was capturing electrical potential due to higher current density. capturing electrical potential due to higher current density.

Figure 7. Comparison the electric field effect and surface charge density of the electrodes using GN-based (a) and conventional Ag/AgCl based electrode (b).

Figure7. 7.Comparison Comparison the the electric electric field field effect using Figure effect and and surface surfacecharge chargedensity densityofofthe theelectrodes electrodes using GN-based andconventional conventionalAg/AgCl Ag/AgCl based GN-based (a)(a)and basedelectrode electrode(b). (b).

Nanomaterials 2016, 6, 156 Nanomaterials 2016, 6, 156

9 of 16 9 of 16

3.2. Measurement of Skin-Electrode Contact Impedances

Skin–electrode contact been of of interest duedue to proving the contact impedance impedancemeasurement measurementhas hasalways always been interest to proving reliability of the collected biopotential. Skin conductivity varies the reliability of the collected biopotential. Skin conductivity variesaccording accordingtotovariations variations of of the conditions of either the stratum corneum (whether hair is on skin or not) or sweat proportions. In of either the stratum corneum (whether hair is on skin or not) or sweat proportions. order to to getget a high-quality signal acquisition with minimal noise, thethe impedance measurement of In order a high-quality signal acquisition with minimal noise, impedance measurement skin–electrode should bebe small and of skin–electrode should small andstable. stable.To Tocharacterize characterizethe theimpedance impedance of of the graphene-coated electrode, aameasurement setup waswas constructed basedbased on earlier techniques reportedreported in the literature measurement setup constructed on earlier techniques in the [4,5]. Here, we Here, measured the electrode–skin contact contact impedance of threeofGN-coated electrodes with literature [4,5]. we measured the electrode–skin impedance three GN-coated electrodes different sizes,sizes, and and compared with that of of thethe conventional with different compared with that conventionalAg/AgCl Ag/AgClelectrode electrode(see (seeFigure Figure 8). Both the conventional conventional Ag/AgCl Ag/AgCl electrodes electrodes and andthe theproposed proposedGN-coated GN-coatedelectrodes electrodes were placed adjacent were placed adjacent to to each other a person’s forearm (between the and wrist the elbow), and both electrodes were each other on aon person’s forearm (between the wrist theand elbow), and both electrodes were attached attached so as to the maintain the same cm). Each measurement wasjust carried just after so as to maintain same distance (3 distance cm). Each(3measurement was carried out after out attaching the attaching the electrodes period 45 s) and aftermoisture. removalThe of skin moisture. The impedance electrodes (in period of 45(in s) and afterofremoval of skin impedance measurements were measurements recorded frequency of 20 to Hzthese to 1measurements, kHz. According these recorded in the were frequency rangeinof the 20 Hz to 1 kHz.range According the to trend of measurements, the trend of impedances with varying frequencies was improved by graphene coating impedances with varying frequencies was improved by graphene coating when compared to that of when compared to that[4] of Baek’s and Meng’s [4]the results. Figurevalues 8 shows impedanceAg/AgCl values of Baek’s [5] and Meng’s results.[5] Figure 8 shows impedance ofthe conventional conventional Ag/AgCl electrode ranges from 445.05 20 Hz) to 13.82 kΩ (at 1tokHz), are electrode ranges from 445.05 kΩ (at 20 Hz) to 13.82 kΩkΩ (at(at 1 kHz), which are similar resultswhich reported similar to results reported in the literature [4,5], and the impedance of the graphene-coated electrode in the literature [4,5], and the impedance of the graphene-coated electrode varies from 65.82 kΩ varies from kΩ (at (at 20 Hz) toThe 5.10results kΩ (atshow 1 kHz). show that the graphene-coated (at 20 Hz) to65.82 5.10 kΩ 1 kHz). thatThe theresults graphene-coated electrode has lower electrode has lower skin–electrode contact impedance compared to the conventional Ag/AgCl skin–electrode contact impedance compared to the conventional Ag/AgCl electrode, resulting in less electrode, resulting in less noise and a higher quality ECG signal, which is shown in Figure 9. Another noise and a higher quality ECG signal, which is shown in Figure 9. Another interesting point from interesting point from Figure 8 is that decreases the contact decreases as the size of(as thementioned electrode Figure 8 is that the contact impedance asimpedance the size of the electrode increases increases (as mentioned earlier, Ambu bigger type electrode is slightly bigger than Covidien type), which earlier, Ambu type electrode is slightly than Covidien type), which results from the increased results from the increased contact area contact area between the electrode and between the skin. the electrode and the skin.

Figure 8. The and Covidien (smaller) types of Theskin–electrode skin–electrodecontact contactimpedance impedanceofofAmbu Ambu(bigger) (bigger) and Covidien (smaller) types conventional Ag/AgCl electrode (shown of conventional Ag/AgCl electrode (shownasasImpedanceAmbu ImpedanceAmbu and and ImpedanceCovidien) ImpedanceCovidien) and ImpedanceGNAmbu and and ImpedanceGNCovidien). ImpedanceGNCovidien). graphene-coated electrodes (shown as ImpedanceGNAmbu

3.3. Measurements of ECG ECG Signals Signals 3.3. Measurements of Electrocardiogram (ECG) using two two types types (Ambu (Ambu and and Covidien) Covidien) of of Electrocardiogram (ECG) signals signals were were recorded recorded using conventional Ag/AgCl electrodes and the newly developed GN-based electrodes at the lead I conventional Ag/AgCl electrodes and the newly developed GN-based electrodes at the lead I position. position. The typical ECG signals from each electrode are shown in Figure 9. In comparison with the

Nanomaterials 2016, 6, 156

Nanomaterials 2016, 6, 156

10 of 16

10 of 16

The typical ECG signals from each electrode are shown in Figure 9. In comparison with the ECG signals obtained with the conventional andAg/AgCl GN-based is observed there was ECG signals obtained with theAg/AgCl conventional andelectrodes, GN-baseditelectrodes, it isthat observed thatno significant between the two signals, thatbut theonly waveforms of the ECGofsignal from there wasdifference no significant difference between the but twoonly signals, that the waveforms the ECG GN-coated electrode were much more stable than conventional electrodes. The P-wave, QRS-complex signal from GN-coated electrode were much more stable than conventional electrodes. The P-wave, (QRS-complex the combination the Q-wave, R-wave andR-wave S-waveand andS-wave represents ventricular QRS-complexis(QRS-complex is theofcombination of the Q-wave, and represents depolarization), and the T-wave clearly visible in both waveforms both Ambu and Covidien ventricular depolarization), andwere the T-wave were clearly visible in bothand waveforms and both Ambu types electrodes. It of is also shownItthat electrodes withelectrodes larger sizes exhibit dueless to a larger andof Covidien types electrodes. is also shown that with largerless sizesnoise exhibit noise due to a larger skin-electrode area. Covidien The GN-based type (smaller electrode skin-electrode contact area. Thecontact GN-based type Covidien (smaller size) electrodesize) provided ECG provided ECG signals that are comparable tosize) Ambu type (larger size) conventional signals that are comparable to Ambu type (larger conventional Ag/AgCl electrode. InAg/AgCl particular, particular, results of electrode both typesdemonstrated of GN-coated the electrode demonstrated theelectrodes potential in theelectrode. results ofInboth types ofthe GN-coated potential use of these use of these electrodes in clinical settings. clinical settings.

Figure Chest-basedECG ECGsignals signalsfrom from the the Ag/AgCl Ag/AgCl electrode electrode Figure 9. 9. Chest-based electrodeand andfrom fromthe theGN-based GN-based electrode using Ambu (largersize) size)and andCovidien Covidien(smaller (smaller size) size) types of using Ambu (larger of electrodes electrodes(a–d). (a–d).

To further evaluate and compare the performance of the GN-coated and Ag/AgCl electrodes, To further evaluate and compare the performance of the GN-coated and Ag/AgCl electrodes, the ear-lead electrode position was suggested to put into the system to see whether the signals are the ear-lead electrode position was suggested to put into the system to see whether the signals are still

Nanomaterials 2016, 6, 156

11 of 16

Nanomaterials 2016, 6, 156

11 of 16

stable,still duestable, to easy fortouse and for ECG for setup and monitoring. Normally, it is quite due easy forsimplicity use and simplicity ECG setup and monitoring. Normally, it isdifficult quite to measure an ECG signal near the ear due to weak surface conduction when compared to lead I position, difficult to measure an ECG signal near the ear due to weak surface conduction when compared to lead I position, and the movementeffect has aon significant on However, the ECG signal. However,electrodes GNand the body movement hasbody a significant the ECGeffect signal. GN-enabled enabled aelectrodes have a great effect on coupling ECG signals from behind-the-ear location to have shown great effect onshown coupling ECG signals from behind-the-ear location when compared whenelectrodes. compared toHere, Ag/AgCl Here, of theafirst electrode of a three-electrode setupbehind was placed Ag/AgCl theelectrodes. first electrode three-electrode setup was placed the ear; behind the ear; the second electrode was attached on the upper neck area, and the last electrode, the second electrode was attached on the upper neck area, and the last electrode, called reference called reference electrode, was placed on the left waist. Again, we also examined the effect of the size electrode, was placed on the left waist. Again, we also examined the effect of the size of electrodes, of electrodes, as Ambu and Covidien types of electrodes were used. Figure 10a,c illustrates the ECG as Ambu and Covidien types of electrodes were used. Figure 10a,c illustrates the ECG signals of the signals of the Ag/AgCl electrode and of the GN-enabled electrode using Ambu-type electrodes Ag/AgCl electrode andear. of Likewise, the GN-enabled electrodethe using Ambu-type electrodes behind located behind the we also measured signals of Ag/AgCl electrode located and of the GN- the ear. Likewise, we also measured the signals of Ag/AgCl electrode and of the GN-coated electrode coated electrode using the Covidien type of electrode, and ECG waveforms are shown in usingFigure the Covidien type of electrode, and ECG waveforms are shown in Figure 10b,d, respectively. 10b,d, respectively.

Figure 10. Ear-lead based ECG signals fromthe theAg/AgCl Ag/AgCl electrode from thethe GN-based electrode Figure 10. Ear-lead based ECG signals from electrodeand and from GN-based electrode using Ambu (larger size) and Covidien (smaller size) types of electrodes (a–d). using Ambu (larger size) and Covidien (smaller size) types of electrodes (a–d).

As can be seen from the figure, it was discovered that there are no significant changes on the

As cansignals be seen fromAmbu the figure, it was discovered no significant changes onwere the ECG ECG using type of electrodes, but onthat thethere otherare hand, acquired ECG signals distorted in the case Covidien type of electrode, we acquired could not ECG identify the P-were and T-waves signals using Ambu typeofofthe electrodes, but on the otherand hand, signals distorted in from (see Figuretype 10c).of However, we and werewe able to observe the P-wave, QRS-complex andfrom the T-there the case ofthere the Covidien electrode, could not identify the Pand T-waves wave within Covidien type GN-based electrode as shown in Figure 10d. Moreover, due to increased (see Figure 10c). However, we were able to observe the P-wave, QRS-complex and the T-wave within contact area with the larger size ofas electrode type10d. of electrode), thedue quality of ECG waveforms Covidien type GN-based electrode shown (Ambu in Figure Moreover, to increased contact area was better than that of the Covidien type of electrode when Figure 10a–c is compared to Figure 10b–d. with the larger size of electrode (Ambu type of electrode), the quality of ECG waveforms was better Furthermore, we evaluated the performance of the electrodes, and ECG signals were analyzed than that of the Covidien type of electrode when Figure 10a–c is compared to Figure 10b–d. to calculate signal-to-noise ratio (SNR) using the following equation [14]: Furthermore, we evaluated the performance of the electrodes, and ECG signals were analyzed to SNR = 20 log(S/(S’ − S)) (1) calculate signal-to-noise ratio (SNR) using the following equation [14]: SNR = 20 log(S/(S0 − S))

(1)

where S is the filtered ECG signal with a frequency ranging from 0.5 Hz to 100 Hz, and S’ is defined as ECG signal without filtering. Before calculation, the power line interference (50 Hz) was removed

Nanomaterials 2016, 6, 156 Nanomaterials 2016, 6, 156

12 of 16

12 of 16

where S is the filtered ECG signal with a frequency ranging from 0.5 Hz to 100 Hz, and S’ is defined from both signals. Table 2 summarizes the SNR of eight different ECG results, which are represented as ECG signal without filtering. Before calculation, the power line interference (50 Hz) was removed in Figures 9 and 10, between a–d, respectively. from both signals. Table 2 summarizes the SNR of eight different ECG results, which are represented in Figures 9 and 10, between a–d, respectively. Table 2. Signal-to-noise ratios (SNRs) of electrocardiography (ECG) signals obtained with different electrodes and placements. Table 2. Signal-to-noise ratios (SNRs) of electrocardiography (ECG) signals obtained with different electrodes and placements.

Electrode Placement

Ambu Type Electrode

Covidien Type Electrode

Electrode Placement

Ambu Type Electrode CovidienAg/AgCl Type Electrode GN-based Ag/AgCl GN-based Chest-lead 27.03 25.21 25.32 GN-based Ag/AgCl GN-based 21.23 Ag/AgCl 17.27 21.7425.32 15.34 Chest-leadEar-lead 27.0322.96 25.21 21.23 Ear-lead 22.96 17.27 21.74 15.34 Results clearly revealed that the proposed GN-based electrode within both Ambu and Covidien Results clearly revealed thequality proposed GN-based electrode both Ambu and Covidien type electrodes provide better that signal and performance thanwithin traditional Ag/AgCl electrodes. type electrodes provide better signal quality and performance than traditional Ag/AgCl electrodes. Another point from the results is that GN-based electrode presents a comparable SNR and ECG signal Another point from the results is that GN-based electrodefor presents a comparable SNR healthcare. and ECG signal even from behind the ear, which is a quite useful location the patients in wearable These even from behind the ear, which is a quite useful location for the patients in wearable healthcare. results also pointed out that SNRs agreed with the results taken from ECG signals (Figures 9 and 10), These results also pointed out that SNRs agreed with the results taken from ECG signals (Figures 9 skin–electrode contact impedance (Figure 8), simulation in Comsol (Figure 7), and the electrical and 10), skin–electrode contact impedance (Figure 8), simulation in Comsol (Figure 7), and the conductivity values to realise the effect of GN. Meanwhile, we have also compared acquired ear-lead electrical conductivity values to realise the effect of GN. Meanwhile, we have also compared acquired ECG signals by the proposed GN-based electrode with recently published work by Da He [28] who ear-lead ECG signals by the proposed GN-based electrode with recently published work by Da He [28] reported ECG signals obtained by Ag/AgCl gel electrodes using a behind-the-ear device. In the results who reported ECG signals obtained by Ag/AgCl gel electrodes using a behind-the-ear device. In the reported, only R-waves appeared clearly including various noisy data, however, our results clearly results reported, only R-waves appeared clearly including various noisy data, however, our results identify QRS-complex, and T-waves with less noise Figure 11). The point clearlyP-waves, identify P-waves, QRS-complex, and T-waves with less (see noisein(see in Figure 11). critical The critical here is that Da He used two active electrodes because of the limited skin area near the ear, thus DRL point here is that Da He used two active electrodes because of the limited skin area near the ear, thus (reference) electrode was omitted in his experiments. DRL (reference) electrode was omitted in his experiments.

Figure Comparison ECGsignals signalsrecorded recorded by by different different electrodes. byby Figure 11.11. Comparison ofofECG electrodes.(a) (a)ECG ECGsignals signalsrecorded recorded Ag/AgCl electrode in Da He’s report using ear-lead position R-peaks are visible); (b) signals ECG Ag/AgCl electrode in Da He’s report using ear-lead position (only(only R-peaks are visible); (b) ECG signals recorded by GN-based electrode using ear-lead position (all P-QRS-T morphology is identified). recorded by GN-based electrode using ear-lead position (all P-QRS-T morphology is identified).

Nanomaterials 2016, 6, 156 Nanomaterials 2016, 6, 156

13 of 16 13 of 16

addition, we we also also evaluated evaluated and and compared compared the theperformance performance of ofthe theGN-based GN-basedand andAg/AgCl Ag/AgCl In addition, electrodes by giving a power spectrum of the recorded ECG signals using the built-in KST (KST is a data plotting and analysis program) function [29] with a Hamming window. Figure 12a,b show the the ECG ECG signals signals from from the the both both Ag/AgCl Ag/AgCl and GN-based electrodes, power spectral density (PSD) of the respectively. It It is is clearly seen that frequency response response curves for the GN-based electrodes illustrate a respectively. better power spectrum spectrum than than Ag/AgCl Ag/AgCl electrodes electrodes where the critical P-QRS-T morphology of the ECG captured. Analysis of the PSD also confirms that the low in the 0–40 Hz frequency range is accurately captured. frequency fluctuations occur for the Ag/AgCl frequency fluctuations occur for the Ag/AgCl electrodes in the 2–5 Hz range, however the noise signal observedafter after40 40Hz Hzfrequency frequencyrange range GN-based electrode. noise signal is acceptable is observed forfor thethe GN-based electrode. ThisThis noise signal is acceptable and and caused by low veryfluctuations low fluctuations during ECG recording. trend ofvarying PSD with varying caused by very during the ECGthe recording. The trendThe of PSD with frequencies frequencies similar to that of Yapici’s results [30]. According to theirthe results, PSD of ECG was similar was to that of Yapici’s PSD resultsPSD [30]. According to their results, PSD the of ECG signals signals showed noisewas signal was minimized using Welch periodogram filteringwhich method showed that the that noisethe signal minimized using Welch periodogram filtering method is whichinto is built intoYapici Matlab. Yapici et al. [30] investigated of GN for afabricating a textile built Matlab. et al. [30] investigated the effect ofthe GNeffect for fabricating textile electrode in electrode in an ECG recording system and they the characterization of electrode frequency an ECG recording system and they analysed theanalysed characterization of electrode frequency response for response for further evaluation. further evaluation.

response of of filtered ECG signals from the the Ag/AgCl and GNFigure 12. Comparison Comparisonofofthe thefrequency frequency response filtered ECG signals from Ag/AgCl and based electrode up to (a) Power spectral density (PSD) of ECG recordings (from Figure 8a) 8a) for GN-based electrode up75toHz. 75 Hz. (a) Power spectral density (PSD) of ECG recordings (from Figure the the Ag/AgCl electrode; (b) PSD of ECG recordings (from Figure 8c) for electrode. for Ag/AgCl electrode; (b) PSD of ECG recordings (from Figure 8c) GN-based for GN-based electrode.

To investigate investigate the the stability stability and and reliability reliability of of the the GN-based GN-based electrode electrode in in long-term long-term monitoring, monitoring, To we continued the experiment using our GN-based electrode on 10 different subjects for we continued the experiment using our GN-based electrode on 10 different subjects for 22 months. months. Although same same GN-based GN-based electrodes electrodes were were used, used, we we were were able able to to clearly clearly observe observe aa typical typical ECG ECG Although waveform including includingP-waves, P-waves,QRS-complex QRS-complex and T-waves. Each time, the electrodes cleaned waveform and T-waves. Each time, the electrodes werewere cleaned with with acetone and prepared to be tested with the next subject. After experiments with testing GNacetone and prepared to be tested with the next subject. After experiments with testing GN-based based electrodes on 10 different subjects, we measured the resistance of the electrode to examine if electrodes on 10 different subjects, we measured the resistance of the electrode to examine if such such residues or microcracks the electrical properties of electrode. the electrode. electrical resistances residues or microcracks affectaffect the electrical properties of the TheThe electrical resistances of of the three electrodes were measured, and they were 0.076 ± 0.019, 0.064 ± 0.024, and 0.089 0.129 Ω, Ω, the three electrodes were measured, and they were 0.076 ± 0.019, 0.064 ± 0.024, and 0.089 ±± 0.129 respectively. Even Even though though aa small small increase increase of of resistance resistance was was discovered, discovered, the the electrode electrode still still had had good good respectively. electrical conductivity with 416 S/m after 2 months of experiments. These experimental results electrical conductivity with 416 S/m after 2 months of experiments. These experimental results showed showed that the GN-enabled is for wearable for more thanand 2 months that this that the GN-enabled electrode electrode is wearable more than 2 months that thisand electrode canelectrode be used can be used on different patients for use in long-term ECG monitoring. on different patients for use in long-term ECG monitoring.

Nanomaterials 2016, 6, 156

14 of 16

4. Conclusions In this paper, a novel graphene (GN)-coated ECG electrode was developed and its performance was tested in terms of quality-of-signal and durability. The electrodes were obtained by CVD grown on Cu method and the structures were transferred through Ag substrates. The experimental results clearly showed improved performance with graphene-coated electrodes. The signal-to-noise ratio has improved by 8%, significantly due to GN coating; the shape of the signal has been much better with GN-coated electrodes than that of electrodes without GN coating. Thus, ECG signal had less low-frequency fluctuation and high-frequency noise with GN-coated versus uncoated. It was also found that quality of GN-coated electrodes was not significantly degraded even after multiple uses (10 times). The characteristics of graphene coating on the conventional electrode was also investigated experimentally using SEM images, Raman spectroscopy, and impedance measurements. The measurements/observations from these experiments have evidently showed the improvement due to graphene coating. The SEM images show the smoother surface of GN coating, which would increase skin-to-electrode contact. The Raman spectroscopy measurement confirms that there is no presence of the 2D-band before GN coating process, however, Raman spectrum consists of a pronounced 2D-peak of comparable intensity to the G-peak and a pronounced reduction in the full width at half maximum (FWHM) of the D- and G-bands (see in Figure 3a,b). As mentioned earlier, these Raman spectroscopy measurements fit with single-layer GN deposition onto the substrate, evidencing separation of itself from the case of graphite or multiple-layer GN deposition. Impedance measurements of the ECG electrodes have shown reduced impedance due to the GN coating compared to that of electrodes without coating, which would increase the sensitivity of the electrode. A finite element modelling (FEM) of skin–electrode was also developed to understand the electrical activities of skin–electrode contact. The simulation results also suggest that the GN coating has improved the current density and electric field in the region of interest, where electrode connects to the skin. These results hold promise for further development of the new nanomaterial-enabled dry electrodes for electrophysiological sensing in wearable technologies. Therefore, the GN-based electrode reflects potential application in not only cardiac activity ECG monitoring systems, but also muscular (EMG) and neural activity (EEG). Supplementary Materials: The following are available online at www.mdpi.com/2079-4991/6/9/156/s1. Figure S1: Raman analysis of bare Ag/AgCl electrode for detecting 2D band in 10 measurements, Figure S2: Raman analysis of Graphene-coated electrode for detecting 2D band in 10 measurements. Although weak peaks occurred during tests, 7/10 peaks presented in 2D band appearances. Acknowledgments: This research was supported by Ministry of National Education of Turkey, the research group of Doclab in Brunel University London, and 2-DTech company based in Manchester. The data used in this publication can be found in doi:10.17633/rd.brunel.3574341. Author Contributions: The research aims and methods were proposed by Numan Celik, Nadarajah Manivannan and Wamadeva Balachandran. The introduction and related works sections were written and reviewed by Numan Celik and Wamadeva Balachandran. The sections on experiments and results were handled by Numan Celik, Nadarajah Manivannan, Andrew Strudwick and Wamadeva Balachandran. Conflicts of Interest: The authors declare no conflict of interest.

References 1.

2. 3. 4.

Yoo, J.; Yan, L.; Lee, S.; Kim, H.; Yoo, H.J. A wearable ECG acquisition system with compact planar-fashionable circuit board-based shirt. IEEE Trans. Inf. Technol. Biomed. Actions 2009, 13, 897–902. [CrossRef] [PubMed] Webster, J.G.; Clark, J.W. Medical Instrumentation; Wiley: New York, NY, USA, 1998. Oh, T.I.; Yoon, S.; Kim, T.E.; Wi, H.; Kim, K.J.; Woo, E.J.; Sadleir, R.J. Nanofiber web textile dry electrodes for long-term biopotential recording. IEEE Trans. Biomed. Circuits Syst. 2013, 7, 204–211. [PubMed] Meng, Y.; Li, Z.; Chen, J. A flexible dry electrode based on APTES-anchored PDMS substrate for portable ECG acquisition system. Microsyst. Technol. 2016, 22, 2027–2034. [CrossRef]

Nanomaterials 2016, 6, 156

5. 6.

7. 8. 9. 10. 11.

12. 13. 14. 15. 16. 17. 18.

19.

20. 21.

22.

23.

24. 25. 26.

27.

15 of 16

Baek, J.Y.; An, J.H.; Choi, J.M.; Park, K.S.; Lee, S.H. Flexible polymeric dry electrodes for the long-term monitoring of ECG. Sens. Actuators A 2008, 143, 423–429. [CrossRef] Hoffmann, K.P.; Ruff, R. Flexible dry surface-electrodes for ECG long-term monitoring. In Proceedings of the 29th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Chicago, IL, USA, 23–26 August 2007; pp. 5739–5742. Baek, H.J.; Lee, H.B.; Kim, J.S.; Choi, J.M.; Kim, K.K.; Park, K.S. Nonintrusive biological signal monitoring in a car to evaluate a driver’s stress and health state. Telemed. e-Health 2009, 15, 182–189. [CrossRef] [PubMed] Lee, S.M.; Sim, K.S.; Kim, K.K.; Lim, Y.G.; Park, K.S. Thin and flexible active electrodes with shield for capacitive electrocardiogram measurement. Med. Biol. Eng. Comput. 2010, 48, 447–457. [CrossRef] [PubMed] Searle, A.; Kirkup, L. A direct comparison of wet, dry and insulating bioelectric recording electrodes. Physiol. Meas. 2000, 21. [CrossRef] Myers, A.C.; Huang, H.; Zhu, Y. Wearable silver nanowire dry electrodes for electrophysiological sensing. RSC Adv. 2015, 5, 11627–11632. [CrossRef] Jung, H.C.; Moon, J.H.; Baek, D.H.; Lee, J.H.; Choi, Y.Y.; Hong, J.S.; Lee, S.H. CNT/PDMS composite flexible dry electrodesfor long-term ECG monitoring. IEEE Trans. Biomed. Eng. 2012, 59, 1472–1479. [CrossRef] [PubMed] Watanabe, M.; Mizukami, K. Well-ordered wrinkling patterns on chemically oxidized poly (dimethylsiloxane) surfaces. Macromolecules 2012, 45, 7128–7134. [CrossRef] Lee, S.K.; Kim, B.H.; Yoo, H.J. Planar fashionable circuit board technology and its applications. J. Semicond. Technol. Sci. 2009, 9, 174–180. [CrossRef] Thompson, C.S.; Abate, A.R. Adhesive-based bonding technique for PDMS microfluidic devices. Lab Chip 2013, 13, 632–635. [CrossRef] [PubMed] Celik, N.; Balachandran, W.; Manivannan, N. Graphene-based biosensors: Methods, analysis and future perspectives. IET Circuits Devices Syst. 2015, 9, 434–445. [CrossRef] 2-DTech LTD. Available online: http://www.2-DTech.com (accessed on 14 April 2016). Mattevi, C.; Kim, H.; Chowalla, M. A Review of Chemical Vapor Deposition of Graphene on Copper. J. Mater. Chem. 2011, 21, 3324–3334. [CrossRef] Li, X.; Cai, W.; An, J.; Kim, S.; Nah, J.; Yang, D.; Piner, R.; Velamakanni, A.; Jung, I.; Tutuc, E.; et al. Large-Area Synthesis of High-Quality and Uniform Graphene Films on Copper Foils. Science 2009, 324, 1312–1314. [CrossRef] [PubMed] Kidambi, P.R.; Ducati, C.; Dlubak, B.; Gardiner, D.; Weatherup, R.S.; Martin, M.-B.; Seneor, P.; Coles, H.; Hofmann, S. The Parameter Space of Graphene Chemical Vapor Deposition on Polycrystalline Cu. J. Phys. Chem. C 2012, 116, 22492–22501. [CrossRef] Reina, A.; Son, H.; Jiao, L.; Fan, B.; Dresselhaus, M.S.; Liu, Z.; Kong, J. Transferring and Identification of Single-and Few-Layer Graphene on Arbitrary Substrates. J. Phys. Chem. C 2008, 112, 17741–17744. [CrossRef] Strudwick, A.J.; Weber, N.E.; Schwab, M.G.; Kettner, M.; Weitz, R.T.; Wünsch, J.R.; Müllen, K.; Sachdev, H. Chemical Vapor Deposition of High Quality Graphene Films from Carbon Dioxide Atmospheres. ACS Nano 2014, 9, 31–42. [CrossRef] [PubMed] Li, X.; Magnuson, C.W.; Venugopal, A.; Tromp, R.M.; Hannon, J.B.; Vogel, E.M.; Vogel, E.M.; Colombo, L.; Ruoff, R.S. Large-area graphene single crystals grown by low-pressure chemical vapor deposition of methane on copper. J. Am. Chem. Soc. 2011, 133, 2816–2819. [CrossRef] [PubMed] Ferrari, A.C.; Meyer, J.C.; Scardaci, V.; Casiraghi, C.; Lazzeri, M.; Mauri, F.; Piscanec, S.; Jiang, D.; Novoselov, K.S.; Roth, S.; et al. Raman spectrum of graphene and graphene layers. Phys. Rev. Lett. 2006, 97. [CrossRef] [PubMed] Thomsen, C.; Reich, S. Double resonant Raman scattering in graphite. Phys. Rev. Lett. 2000, 85. [CrossRef] [PubMed] Ferrari, A.C. Raman spectroscopy of graphene and graphite: Disorder, electron–phonon coupling, doping and nonadiabatic effects. Solid State Commun. 2007, 143, 47–57. [CrossRef] Chen, Z.; Ren, W.; Gao, L.; Liu, B.; Pei, S.; Cheng, H.M. Three-dimensional flexible and conductive interconnected graphene networks grown by chemical vapour deposition. Nat. Mater. 2011, 10, 424–428. [CrossRef] [PubMed] Umer, D.; Ellahi, G.; Ellahi, W.; Ayub, S.; Din, Z.M.U. Bio-electrical Impedance Plethysmography-Designing an Efficient Non-invasive Electrode for Cardiac Rate Analysis. Int. J. Endorsing Health Sci. 2013, 1, 29–32.

Nanomaterials 2016, 6, 156

28. 29. 30.

16 of 16

He, D.D.; Winokur, E.S.; Sodini, C.G. An Ear-Worn Vital Signs Monitor. IEEE Trans. Biomed. Eng. 2015, 62, 2547–2552. [CrossRef] [PubMed] KST Visualizing Data. Available online: https://kst-plot.kde.org/kst1/handbook/plugins-vector-processing. html (accessed on 12 March 2016). Yapici, M.K.; Alkhidir, T.; Samad, Y.A.; Liao, K. Graphene-clad textile electrodes for electrocardiogram monitoring. Sens. Actuators B 2015, 221, 1469–1474. [CrossRef] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).

Graphene-Enabled Electrodes for Electrocardiogram Monitoring.

The unique parameters of graphene (GN)-notably its considerable electron mobility, high surface area, and electrical conductivity-are bringing extensi...
5MB Sizes 1 Downloads 12 Views