Stretchable and Biocompatible Transparent Electrodes for Multimodal Biosignal Sensing from Exposed Skin

Real‐time health monitoring technology in daily life requires mechanically robust and transparent electrodes for multimodal biosignal sensing from exposed human epidermis. Here, highly stretchable transparent electrodes comprising a water‐dispersed conductive polymer, poly(3,4‐ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS), and a protic ionic liquid (IL), 3‐methylimidazolium:bis(trifluoromethylsulfonyl)amide (p‐MIM:TFSI) are reported. Owing to the high water miscibility of p‐MIM:TFSI and its favorable ion exchange capability with PEDOT:PSS, PEDOT:PSS/p‐MIM:TFSI transparent electrodes show enhanced electrical conductivity (σ = 450 S cm−1) and thin‐film stretchability represented by crack onset strain (εc) exceeding 50%. These electrodes outperform other PEDOT:PSS electrodes processed with an aprotic counterpart, 1‐ethyl‐3‐methylimidazolium(EMIM):TFSI, or a traditional ionic salt, Li:TFSI. The PEDOT:PSS/p‐MIM:TFSI thin‐film electrodes are also biocompatible and conformally adhere to human skin; therefore, multimodal biosignals including electrocardiogram, electrooculogram, and electromyogram with high signal‐to‐noise ratios from exposed epidermis on human faces and arms under various measurement conditions mimicking daily activities are collected. Considering the importance of light penetration through human skin for stable biological activity during biosignal monitoring, the results can broaden the applicability of daily‐use wearable biosignal sensors by applying them to exposed human skin.


Electrical and Optical Properties
Recently, we developed highly conductive and stretchable PEDOT:PSS thin films using a new p-IL based on 3-methylimidazolium and bis(trifluoromethylsulfonyl)amide as cation and anion (p-MIM:TFSI), respectively (Figure 1a). [40] In this study, we investigated the films' usefulness as stretchable transparent electrodes for real-time biosignal monitoring by comparing them with PEDOT:PSS thin films prepared using an aprotic IL (ap-IL), 1-ethyl-3-methylimidazolium(EMIM):TFSI, or an ionic salt, Li:TFSI. Note that we chose TFSI to avoid any undesired crystallization of ILs or rough surface morphology that could have hindered conformal contact with human skin. [41] As transparent electrodes, PEDOT:PSS/ionic additive thin films exhibited high average %T over 95% in wavelengths of 300-800 nm because the films were thin, ≈100 nm (Figure 1b; Figure S1, Supporting Information).
Owing to the enhanced water solubility of our p-MIM:TFSI and its enhanced miscibility with PEDOT:PSS, the PEDOT:PSS/p-MIM:TFSI thin films showed smooth surface morphology (Figure 1e). By contrast, the PEDOT:PSS/ EMIM:TFSI thin films showed aggregates on the surface that were more clearly seen in the scanning electron microscopy (SEM) images shown in Figure 1f. Contrary to the smooth surface and cross-section of the PEDOT:PSS/p-MIM:TFSI thinfilm, the PEDOT:PSS/EMIM:TFSI thin-film exhibited many aggregates of ≈10 µm; although the PEDOT:PSS/Li:TFSI thinfilm showed no aggregates, it exhibited granular morphology in the cross-sectional view, which could have led to the larger standard deviations for σ and R s (Figure 1c,d). We also used energy-dispersive X-ray (EDX) analysis to investigate and identify the components of the aggregates on the PEDOT:PSS/ EMIM:TFSI thin-film ( Figure 1g). The distributions of nitrogen (N) and fluorine (F) in the EDX images revealed that the aggregates were composed of EMIM:TFSI, presumably owing to the lower water solubility ( Figures S2-S4, Supporting Information).
We then used a universal testing machine (UTM) to compare the mechanical properties of freestanding PEDOT:PSS films prepared with and without ionic additives. Consistent with the finding that ionic additives are known to increase the stretchability of PEDOT:PSS/ionic additive composites, PEDOT:PSS freestanding films prepared with ionic additives showed reduced elastic modulus (E) and tensile strength (S) as well as enhanced fracture strain (ε f ) (Figure 2c,d). Notably, the PEDOT:PSS/p-MIM:TFSI freestanding film showed the lowest E (= 10 MPa) and S (= 5 MPa), 39 and 8 times lower than those of the bare PEDOT:PSS freestanding film (E = 388 MPa; S = 38 MPa) respectively. The E and S of the PEDOT:PSS/p-MIM:TFSI film were also lower than those of the PEDOT:PSS/ EMIM:TFSI and PEDOT:PSS/Li:TFSI films. Furthermore, the PEDOT:PSS/p-MIM:TFSI film yielded the highest ε f ( = 65%) among the freestanding films used in this experiment. Based on the previous study, we attribute the improved stretchability of the PEDOT:PSS/p-MIM:TFSI films to the strong hydrogen bonds in the PEODT + -TFSI --p-MIM + -PSSnetworks formed by favorable ion exchange between PEDOT:PSS and p-MIM:TFSI (Figure 2e). [40,42]

Biocompatibility
To verify the biocompatibility of our new IL (i.e., p-MIM:TFSI), we measured the PEDOT:PSS/p-MIM:TFSI thin films' cytotoxicity by testing the viability of fibroblasts cells that play crucial roles and serve diverse functions in the body. [43] For comparison, we used polyethyleneimine (PEI) and d-sorbitol as positive and negative controls, respectively. [25,44] In addition, we selected dimethylsulfoxide (DMSO), which is known to enhance the electrical conductivity of PEDOT:PSS but is toxic, for PEDOT:PSS/ DMSO thin films as an additional positive control. [45,46] We also investigated the cytotoxicity of bare PEDOT:PSS and SEBS substrates to verify the safety of PEDOT:PSS/p-MIM:TFSI/SEBS electrodes for biosignal sensors.
Cell viabilities of 15% and 99% for PEI and d-sorbitol, respectively, show the accuracy of the cytotoxicity test (Figure 3). Bare PEDOT:PSS thin films and SEBS substrates exhibited no cytotoxicity, with cell viabilities over 90%. However, considering that the borderline of toxicity is ≈80%, [16] PEDOT:PSS/ DMSO thin films showed toxicity based on the reduced cell viability of 63%. Notably, their cell viabilities were under 80% even after eight dilutions because of the high toxicity of DMSO extracted from the PEDOT:PSS/DMSO thin films. [47] In contrast, PEDOT:PSS/p-MIM:TFSI thin films showed remarkably high cell viability over 95%. These results demonstrate the usefulness of our PEDOT:PSS/p-MIM:TFSI thin films prepared on SEBS substrates as stretchable transparent electrodes that can be applied to human epidermis for biosignal monitoring.

Biosignal Monitoring
To validate the applicability of our stretchable and transparent PEDOT:PSS/p-MIM:TFSI electrodes, we performed ECG measurements from the forefingers (Figure 4a), which we chose to show the reliability of our electrodes under various bending conditions that mimicked daily activities. To this end, two PEDOT:PSS/p-MIM:TFSI electrodes were mounted on each forefinger. Next, the other PEDOT:PSS/p-MIM:TFSI electrode was attached on the right inner ankle as a reference electrode. ECG signals, which are composed of P wave, QRS complex, and T wave, [48] were measured as differential signals between the electrodes on the forefingers (Lead I). The signals were then amplified and filtered through a custom-made circuit. The peak-to-peak value of the QRS complex was about 0.5 V in the output signals from the circuit. Because the amplification factor at 10 Hz (major frequency of QRS complex) through the circuit was ≈978, we estimated the actual magnitude of the QRS complex before amplification to be 0.51 mV, which is in the typical range (0.5-3.0 mV depending on the locations of electrodes). [49,50] ECG signals obtained using our PEDOT:PSS/p-MIM:TFSI electrodes showed well-defined PQRST waves (Figure 4b).
Notably, their noise was lower than that obtained using the PEDOT:PSS/EMIM:TFSI and PEDOT:PSS/Li:TFSI electrodes, and comparable to that of commercial electrodes ( Figure S5, Supporting Information). Furthermore, the PEDOT:PSS/ EMIM:TFSI electrodes yielded signals with unclear PQRST waves owing to high background noise. The signals measured using the PEDOT:PSS/Li:TFSI electrodes were less noisy than the signal obtained with PEDOT:PSS/EMIM:TFSI electrodes, but the P waves were recognized less frequently than those in In order to employ the high %T of our PEDOT:PSS/p-MIM:TFSI electrodes, we monitored EOG and EMG signals with the electrodes attached on facial skin (Figure 5a). The circuits used to measure the EOG and EMG signals were similar to the one used for the ECG measurement above, but they had different amplification factors and cut-off frequencies of active filters optimized for EOG and EMG measurements. EOG measures the difference in electrical potential induced upon the bipolarity of the eyeball. To obtain EOG signals, two electrodes were attached near the eyes, and the other electrode was placed at the back of the ear as a reference (Figure 5a). The electrodes were placed diagonally across the eye to enable detecting both vertical and horizontal movements. We obtained the EOG signals during various eye movements that occur frequently (upward, downward, leftward, and rightward; Figure 5b). In contrast with the flat base signal, the EOG signals obtained under the different eye movement conditions and blinking showed distinguishable responses to eye movements in all directions and to blinking. The noise levels of EOG signals obtained with our PEDOT:PSS/p-MIM:TFSI electrodes were similar to that of commercial electrodes ( Figure S8, Supporting Information).
In addition, we monitored EMG signals with PEDOT:PSS/p-MIM:TFSI electrodes attached on the chin. Because EMG detects the differences in electrical potentials generated by muscle fibers, the two electrodes were placed on the chin to detect muscle movement, while the other reference electrode was attached to the back of the ear (Figure 5a). EMG signals with larger magnitudes were clearly observed when the participant's teeth were clinched than those obtained with the flat and low baseline signals (Figure 5c). Assuming a human experiencing rapid eye movement sleep, these EMG signals along with the EOG signals can be used for real-time sleep staging. [51,52] Our electrodes were also applied to a human arm to demonstrate the applicability of the PEDOT:PSS/p-MIM:TFSI electrodes for monitoring EMG signals from various muscle positions, and the EMG signals were properly monitored when the participant grasped a fist repeatedly (Figure 5d,e). Note that the noise levels of EMG signals obtained with our PEDOT:PSS/p-MIM:TFSI electrodes were comparable to those of commercial electrodes ( Figure S9, Supporting Information). These results demonstrate the excellence of our stretchable and transparent PEDOT:PSS/p-MIM:TFSI electrodes compared to the currently reported transparent electrodes (Table S1, Supporting Information).

Conclusion
We have demonstrated highly conductive, stretchable, biocompatible, and transparent PEDOT:PSS electrodes based on a new p-IL. This p-IL made the PEDOT:PSS highly stretchable in thin-film form (≈100 nm in thickness) due to strong hydrogen bonds induced between PEDOT:PSS and p-IL ions; therefore, the PEDOT:PSS/p-IL electrodes enabled the real-time monitoring of multimodal biosignals from exposed human skin (e.g., human face, fingers, and arms). Our results can broaden the applicability of stretchable and transparent organic electrodes for wearable biosignal sensors, thus accelerating the commercialization of various epidermal electronics useful for health monitoring in daily life.

Experimental Section
Preparation of Electrodes: The p-MIM:TFSI and EMIM:TFSI aqueous solutions were prepared by dissolving each IL in deionized water with a concentration of 33.8 mg mL −1 . The solutions were subsequently stirred for 10 min using a vortex. Next, each IL solution was blended with PEDOT:PSS aqueous dispersion with a PEDOT:PSS/IL ratio of 1/1.3 by weight. For the PEDOT:PSS/Li:TFSI solutions, Li:TFSI was directly added into PEDOT:PSS aqueous dispersion with a PEDOT:PSS/ Li:TFSI ratio of 1/0.7 by weight. Then, all PEDOT:PSS/ionic additive solutions were vigorously stirred for over 4 h in a vortex. The solutions were then spin cast on precleaned rigid substrates (glass, quartz, silicon oxide, and silicon) or stretchable SEBS substrates to fabricate www.advelectronicmat.de the additive thin films. Before the PEDOT:PSS/ionic additive solutions were spin cast, the SEBS substrates were treated by UV-ozone for 1 h. The SEBS substrates were spin cast from a toluene solution (a concentration of 300 mg mL −1 ) on glass substrates passivated by a perfluorodecyltrichlorosilane (FDTS) self-assembled monolayer. The SEBS substrates were then annealed at 70 °C for 30 min in ambient air for curing. Finally, PEDOT:PSS/ionic additive thin films were annealed at 140 °C (for rigid substrates) or 70 °C (for SEBS substrates) for 15 min in ambient air, respectively.
UV-Vis-NIR Absorption Spectroscopy: A commercial UV-Vis-NIR spectrophotometer (V770, Jasco Ltd.) was used to measure the transmittance spectra of PEDOT:PSS/ionic additive thin films. The films were prepared on quartz substrates. Then, the transmittance spectra were obtained in the visible spectral range (300-800 nm) with a scan rate of 400 nm min −1 .
Electrical Conductivity: The PEDOT:PSS/ionic additive electrodes were prepared on glass or silicon oxide substrates and mounted on a commercial four-point probe conductivity measurement system (T2001A3-UK, Ossila Inc.) to obtain R s . The film thicknesses (t) were determined using atomic force microscopy (AFM5100N, Hitachi Ltd.) and field-emission SEM (FE-SEM SU8220, Hitachi Ltd). The σ values were then determined from the R s and t using the following equation: Optical Microscopy: The OM images were obtained using an optical microscope (Eclipse LV100N POL, Nikon Inc.) to investigate surface morphology and ε c .
Scanning Electron Microscopy: The PEDOT:PSS/ionic additive thin films prepared on silicon substrates were stored in a vacuum desiccator overnight to remove residual solvents. The top and cross-sectional SEM images were captured at 1 and 10 kV, respectively, using FE-SEM (FE-SEM SU8220, Hitachi Ltd.).
Crack Onset Strain: The PEDOT:PSS/ionic additive thin films prepared on SEBS substrates were mounted in a customized stretching machine and stretched at each tensile strain (ε = 10%, 20%, 30%, 40%, and 50%). Additional OM images were taken after releasing the thin films from applied tensile strain. The ε c values of the PEDOT:PSS/ionic additive thin films were determined by OM images, where cracks start to propagate at ε c under increasing tensile strain.
Stress-Strain Analysis: The freestanding PEDOT:PSS and PEDOT:PSS/ ionic additive thick films were prepared by drop casting the solutions on FDTS-passivated silicon dioxide substrates. The drop-cast films were annealed at 50 °C for over 12 h in ambient air. Then, the freestanding www.advelectronicmat.de films were peeled off from the silicon dioxide substrates and subsequently mounted in a commercial UTM (AGX-500N, Shimadzu Ltd.). The stress-strain curves were measured at a rate of 5 mm min −1 in ambient air.
In Vitro Cytotoxicity Test: To evaluate the in vitro biocompatibility of the films, mouse normal fibroblasts (NIH3T3) were cultured in Dulbecco's modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% antibiotic-antimycotic (AA) in a humidified 5% CO 2 atmosphere at 37 °C. The cytotoxicity of the elution solution from the films was analyzed according to ISO 10993-12 guideline. [53,54] First, the films (2 × 3 cm 2 ) were immersed in 1 mL of cell culture medium without FBS under 100 rpm and 37 °C for 72 h with gentle stirring. The elution solution was collected and diluted 1, 2, 4, and 8 times. Then, the fibroblasts were seeded onto a 96-well plate at a density of 10 000 cells per well and incubated at 37 °C overnight. Then, the cells were treated with the elution solution from the films (SEBS, PEDOT:PSS, PEDOT:PSS/DMSO, and PEDOT:PSS/p-MIM:TFSI), polyethyleneimine (as a positive control, 20 µg mL −1 ), and d-sorbitol (as a negative control, 1 mg mL −1 ). After 24 h, the cell viability was assessed using Cell Counting Kit-8 (CCK-8) assay. [55] All analyses were performed in triplicate.
Biosignal Measurement: The PEDOT:PSS/ionic additive electrodes were prepared on SEBS substrates according to the same fabrication procedure described above. The electrodes were connected to copper flat wires using silver paste and carbon tape as conductive adhesives. The skin beneath the copper wire was protected with insulating tape (Micropore, 3M Ltd.) to avoid undesired signals through copper/skin interfaces. The electrodes and wires were covered with 3M medical adhesive tape (Tegaderm, 1622 W, 3M Ltd.) for stable electrical contact during the measurements. The circuits used for the biosignal measurements were composed of an instrumentational amplifier (IA), two active filters, and a passive filter. The IA was used to measure the potential difference between a pair of electrodes with a gain of 50, and the output signal from the IA was filtered and amplified through a high pass filter (HPF), low pass filter (LPF), and 60 Hz notch filter. The HPF and LPF bandwidths were set to 0.4-150 Hz for ECG measurements, 0.1 -75 Hz for EOG measurements, and 10-130 Hz for EMG measurements. Through the circuits, the signal amplification factors were ≈1000 in 1-10 Hz for ECG, 1100 in 0.2-10 Hz for EOG, and 540 in 90-130 Hz for EMG. The analog output signals from the circuits were measured and recorded through a data acquisition board and a custom-made program based on LabVIEW (both National Instruments Ltd.). These experiments were performed by attaching stretchable transparent electrodes on human skin in compliance with the protocol approved by the Institutional Review Board (IRB) at Ewha Womans University. The volunteer (M.K.) agreed to all tests and data in the manuscript with informed consent.
Impedance Measurement: The impedance spectra were obtained using an electrochemical workstation (VMP3, BioLogic Inc.) from 1 to 10 5 Hz with a voltage of 5 mV. The PEDOT:PSS-based electrode was used as the working electrode (WE), and Ag/AgCl gel electrodes were used as the reference electrode (RE) and counter electrode (CE). The distances from the WE center to the RE and CE were both 4 cm, and the contact area of the WE electrode was about 1 cm 2 .

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.