Water-Resistant and Skin-Adhesive Wearable Electronics Using

Research Article Cite This: ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

www.acsami.org

Water-Resistant and Skin-Adhesive Wearable Electronics Using Graphene Fabric Sensor with Octopus-Inspired Microsuckers Sungwoo Chun,†,‡ Wonkyeong Son,§ Da Wan Kim,‡ Jihyun Lee,† Hyeongho Min,† Hachul Jung,∥ Dahye Kwon,∥ A-Hee Kim,∥ Young-Jin Kim,∥ Sang Kyoo Lim,§ Changhyun Pang,*,†,‡ and Changsoon Choi*,§ ACS Appl. Mater. Interfaces Downloaded from pubs.acs.org by UNIV OF LOUISIANA AT LAFAYETTE on 04/29/19. For personal use only.

†

Department SKKU Advanced Institute of Nanotechnology (SAINT) and ‡School of Chemical Engineering, Sungkyunkwan University (SKKU), Suwon-si, Gyeonggi-do 16419, Republic of Korea § Department of Smart Textile Convergence Research, Daegu Gyeongbuk Institute of Science and Technology (DGIST), Daegu 42988, Republic of Korea ∥ Medical Device Development Center, Osong Medical Innovation Foundation, Osong, Cheongju-si 28160, Republic of Korea S Supporting Information *

ABSTRACT: Wearable and skin-attachable electronics with portable/wearable and stretchable smart sensors are essential for health-care monitoring devices or systems. The property of adhesion to the skin in both dry and wet environments is strongly required for efficient monitoring of various human activities. We report here a facile, low-cost, scalable fabrication method for skin-adhesive graphene-coated fabric (GCF) sensors that are sensitive and respond fast to applied pressure and strain. With octopus-like patterns formed on the side of the GCF that touches the skin, the GCF adheres strongly to the skin in both dry and wet environments. Using these characteristics, we demonstrate efficient monitoring of a full range of human activities, including human physiological signals such as wrist pulse and electrocardiography (ECG), as well as body motions and speech vibrations. In particular, both measurements of ECG and wrist-bending motions were demonstrated even in wet conditions. Our approach has opened up a new possibility for wearable and skin-adherent electronic fabric sensors working even in wet environments for health-care monitoring and medical applications in vitro and in vivo. KEYWORDS: fabric sensor, pressure sensors, strain sensors, wearable sensors, graphene-coated fabrics

■

such as extraordinary mechanical strength with flexibility, good electrical conductivity, and chemical stability.16,17 In addition to the choice of sensing materials, it is also very important to determine the underlying substrate, particularly for wearable and skin-attachable applications. The substrate plays a critical role in determining electromechanical response characteristics (e.g., response time, deformability, and dynamic response) of the ultimate sensor system. As has been demonstrated recently, the coating of graphene nanoparticles on a stretchable yarn or fabric substrate directly provides the wearable functionality with good sensory properties. Indeed, a number of graphenecoated yarn-based strain or pressure sensors have been reported recently, and they demonstrate effective monitoring of various human activities.8,18−21 However, the yarn-based sensors do not achieve direct conformal contact with the human skin that may be rough and hairy, but it is actually very important to make good contact with the sensors on the skin to measure the biosignals effectively.22,23 Although chemical adhesives have been mostly applied to increase the surface adhesion of sensors, they may often cause skin damages as well

INTRODUCTION Wearable and skin-attachable electronics have recently attracted great attention because of their potential applicability in portable/wearable and stretchable smart electronic devices for health-care monitoring devices or systems.1−5 As an essential component, stretchable and wearable strain or pressure sensors have been particularly identified as suitable for measuring various human activities, including not only human physiological signals such as wrist pulse, heartbeat, and electrocardiography (ECG) but also body motions (e.g., finger, elbow, face, knee, etc.) and subtle skin vibrations.6−11 For reallife usage of these applications, the sensors are required to exhibit a significantly large electrical change with applied strain over a wide sensing range. Furthermore, the sensors should be lightweight, thin, stretchable, simple, and cost-effective fabrication, as well as conformably adherent to the human skin without skin damage after detachment. Much effort has been recently invested in realizing sensors that satisfy these requirements by introducing nanomaterials such as nanoparticles, nanowires, nanofibers, carbon nanotubes (CNT), and graphene as sensing materials.8,10−15 Notably, graphene-nanoparticle-based percolation systems are among the strong candidates because they are based on easy and facile solution processes, as well as having unique material properties © XXXX American Chemical Society

Received: March 8, 2019 Accepted: April 18, 2019

A

DOI: 10.1021/acsami.9b04206 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 1. Fabrication and characterization of GCF sensors. (a) Schematic illustration of the graphene-coated fabric (GCF) fabrication process. The OPs are patterned on the rGO fabric. (b, c) Scanning electron microscope (SEM) images of (b) rGO fabric sample with coating of rGO-PDMS composite for conductive passivation and (c) OP-engraved rGO fabric. (d) The electrical conductivities of the GCF samples with different numbers of dip-coatings of GO (50 mg/mL) for 20 min. (e) Strain−stress curve of GCF sample with continuous tensile strain. Young’s modulus is estimated as ∼35.5 MPa. (f) Contact angles of water droplet for each fabrication process. Insets show optical images of the rGO-coated fabric and GCF sample.

strength between graphene oxide (GO) and the fabric surface.24−26 The plasma-treated fabric was coated with GO sheets dispersed in deionized (DI) water using an easy dipcoating method. The well-dispersed GO sheets were uniformly coated on the surface of the fabric, thus preserving the surface morphology of the pristine fabric perfectly. The interacting force maintaining the adhesion strength is sufficient to withstand external mechanical stresses (see Figure S2). The GO-coated fabric was soaked in a solution of L-ascorbic acid (L-AA)27 to achieve a conductive reduced graphene oxide (rGO) fabric. The reduction process using the L-AA solution is simple, fast, and cost-effective, produces no physical or chemical damage, and is suitable for the reduction of GOcoated fibers (or fabrics). The presence of graphene on the fabric surface was confirmed to be successful through Raman resonance observations, and the result indicated typical multilayered rGO sheet peaks with a G/2D ratio greater than 1 (the G band was at 1586 cm−1 and the 2D band was at 2683 cm−1) including many defects presented by the D band intensity17 (see Figure S3, Supporting Information). The reduction process does not cause a prominent change in morphology except for changing nonconductive GO to conductive rGO, as shown in Figure 1b. In fact, the perfection of the GO coating on the fabric and its reduction could be easily recognized through the change in the color of the fabric (see Figure S1, Supporting Information). The fabric was initially gray but became brown with the GO coating. Furthermore, the reduction process causes the fabric to be dark. The rGO and poly(dimethylsiloxane) (PDMS) mixture solution diluted in hexane was dip-coated on the rGO-coated fabric, followed by the processes of a vacuum treatment (2 h) and baking (2 h at 100 °C). The dilution using hexane that is

as being highly vulnerable to wet environments. Thus, addressing these issues with wearable and skin-adherent yarn-based electronic sensors is very important for realizing future wearable and skin-adhesive electronic sensor devices and systems. In this work, we developed a facile, low-cost, scalable fabricating method for water-resistant and skin-adhesive wearable graphene-coated fabric (GCF) sensors. The GCF sensors with graphene particles are thin, lightweight, highly conductive, and stretchable. The resulting GCF sensor exhibits sensitive responses to statically and dynamically applied strains with low detection limits and has reliable responses to reproducible operations for both lateral and vertical strains. Inspired by octopus-like patterns (OPs) formed on the side of the fabric that touches the skin, the OPs allow strong adhesion forces on the skin surface, which may be dry/wet and rough. Using these characteristics, we demonstrated the monitoring of a full range of human activities on the human skin, including human physiological signals such as wrist pulse and electrocardiography (ECG) as well as body motions and speech vibrations. Simultaneous measurements of ECG and wristbending motion were demonstrated in dry and wet conditions with continuous bending motions, enabled by skin adherence and wet resistance properties.

■

RESULTS AND DISCUSSION Schematic illustrations of the fabrication of the GCF samples with OPs are shown in Figure 1a and Figure S1 (Supporting Information). A fabric composed of an elastic polyurethane and polyester fiber mixture was used for the substrate of the pristine fabric. The fabric was treated with O2 plasma, thus producing reactive groups and radials for improving adhesion B


Research Article


Figure 2. Piezoresistive response of GCF sensors under applied pressure and strain. (a) The piezoresistive response is represented by the rate of resistance change with statically applied pressures of 0.1−500 kPa. The pressure sensitivity (S = (ΔR/R0)/ΔP) is estimated as 0.74% kPa−1 (0−35 kPa) and 0.04% kPa−1 (35−500 kPa) with a linear fit. (b) The piezoresistive response is represented by the rate of resistance change with lateral strains of 0−30%. The gauge factor (GF = (ΔR/R0)/ε) is estimated as 4.13 (R2 = 99) with a linear fit. (c) Minimum pressure and (d) minimum strain detection limits. The GCF sensor can detect the loading−unloading of a deionized (DI) water droplet with a vertical pressure as low as ∼12 Pa and a loading−unloading operation with a lateral strain as low as ∼0.5%. Reproducible operations of (e) 6000 loading−unloading cycles with a vertical pressure of 40 kPa with a frequency of 1.7 Hz and (f) 1000 stretch−release cycles with a tensile strain of 20%.

removed later is critical for coating the viscous PDMS-based solution on the fabric without clogging the surface. The passivated thin rGO and PDMS composite film plays a role not only in preventing the rGO particles from falling off the fabric when an external force is applied but also in forming a conductive path between the rGO fabric and the surface to which it is attached, such as the human skin. As shown in Figure 1c, the OPs were formed on the composite film using a conventional partial filling technique.28 Although the morphology of the fabric is rather faded, the morphology is relatively well maintained after the OPs. The formation of OPs using a thin and conductive composite film on the fabric substrate imparts the wearable functionality with good sensory properties, comparable to an existing patch-type conductive adhesive.28 The as-prepared GCF samples are lightweight (0.015 g/ cm2) and soft enough to induce mechanical deformations such as stretching, bending, twisting, and folding (see Figure S4, Supporting Information). The electrical conductivities of the GCF samples with different numbers of dip-coatings of GO were investigated (Figure 1d). The conductivity increased from 0.068 to 0.216 S/m with the number of dip-coatings from 1 to 4, respectively. We also evaluated the mechanical

properties of the GCF sample using the stress−strain curve for tension. Figure 1e indicates that the rGO fabric could be stretched up to ∼60% mechanically, and its Young’s modulus is ∼35.5 MPa, although the mechanical properties depend strongly on the pristine fabric used (see Figure S5, Supporting Information). The contact angles of the water droplet were investigated for each fabricating process (Figure 1f). The pristine fabric was hydrophilic, whereas the rGO-coated fabric with PDMS coating was hydrophobic with a contact angle of ∼109° because of the surface property of intrinsic graphene, thus enabling self-cleaning and reduction of noise effects by water molecules. The piezoresistive response characteristics of the GCF pressure and strain sensors were investigated (Figure 2). The sensors used for measurements were completed by forming an electrical connection using a thick rGO film on both sides of the GCF samples, followed by passivation with a Cu tape (see Figure S6, Supporting Information). Figure 2a presents the piezoresistive response represented in a resistance change rate by the applied pressure on the GCF sensor. The static pressures are applied using a weight-stacking method for an accurate measurement,29 and the applied pressure range was 100−500,000 kPa, which corresponds to the entire pressure C


Research Article


Figure 3. Adhesive properties of GCF with OPs. (a) Photograph showing high adhesion of GCF sample on the human skin because of octopus-like patterns (OPs). (b) SEM image of OPs formed on one side of the GCF that touches the skin. The OPs show the dimensions (diameter 100 μm, height 100 μm, and spacing 100 μm) in a hexagonal arrangement with optimization. Right images show top-view and cross-sectional-view SEM images of individual OP. (c) Normal adhesion strengths for the GCF samples with and without OPs in dry and underwater conditions on a silicon wafer with a preload of 4 N/cm2. (d) Photograph showing strong wet adhesion (weight 200 g) of GCF sample with OPs in an underwater environment.

detection range of human touch perception.30 The contact area over which the pressure was applied was 1 × 1 cm2. The GCF sensors are fabricated to be conductive with the initial resistance (R0) without application of pressure. The applied pressure increases the resistance (R) because a vertical pressure reduces the electrical conductivity by decreasing the physical contact areas between the rGO sheets because of a dominant positive piezoresistive effect.31 In fact, the pressure sensors based on conductive graphene particles (e.g., rGO and graphene flake) presented a decrease in resistance with applied pressure because pressure enhances the conducting path vertically,32 indicating a negative piezoresistive effect. However, in this case, the passivation process using the rGO and PDMS composite on the rGO fabric places the GCF pressure sensor under a dominantly positive piezoresistive effect that induces the destruction of the existing rGO conductive networks by lateral stretching of the rGO filler material itself. The sensitivity (S) is defined as S = (ΔR/R0)/ΔP, where ΔR is the change in the resistance (R − R0) in response to the change in the applied vertical pressure (ΔP). The measured resistance change rates do not increase linearly with increasing pressure. In the pressure range below 35 kPa, the sensitivity (S1) was estimated to be 0.74% kPa−1. When the vertical pressure increased from 35 to 500 kPa, the sensitivity decreased compared with the values in the lower pressure range. The sensitivity (S2) in this range is estimated to be 0.04% kPa−1 through a linear fit. Such saturation in sensitivity is attributed to the observed nonlinear behavior, indicating a gradual decrease in the resistance change ratio. The GCF pressure sensor also presented highly sensitive and reliable piezoresistive responses to dynamic vertical pressures that were applied with 10 constant and periodic pressure pulses of 50 kPa and also with increasing pressure pulses of 5, 10, 35, 100, 300, and 500 kPa (see Figure S7, Supporting Information). We developed a 6 × 6 pressure sensor array to demonstrate multiarray detection ability with sensor integration. This array can be practically applied to detect the spatial distribution of pressure. A weight of 14.3 g was placed on an

area of 200 mm2, corresponding to a vertical pressure of 0.7 kPa, in the center of the sensor array. The result depicts the two-dimensional (2D) response image from each sensor subjected to the pressure. The contrast was well mapped according to the local pressure distribution in the contact area (see Figure S8, Supporting Information). The lateral strain-sensing properties of the GCF sensors were also confirmed. The relative resistance responses for the applied lateral strains of 0−30% are shown in Figure 2b. Before the strain, the resistance (R0) of the sensors was determined by the natural networks of as-coated rGO particles. The rGO particles overlapped within a certain distance, thus showing an electrical conductivity value depending on the physical interaction between the particles. A change in film morphology by applying strain on the sensor was expected to increase the resistance (R) of the sensor because the electrical interactions between the rGO particles decreased with an increasing distance between the particles. Ultimately, the resulting resistance change (ΔR = R − R0) is caused mostly by the change in the distance of interaction between the rGO particles. For the gauge factor (GF) defined as GF = (ΔR/R0)/ ε, where ΔR is the resistance change and ε is the applied strain, the GF was estimated as 4.13 (0−30%) with a linear fit. Additionally, the GCF sensors could detect extremely small pressures and strains. The sensor detected continuous loadings of DI water droplets as low as ∼12 Pa (Figure 2c). In addition, the sensor could measure a loading-and-unloading operation of lateral strain as low as ∼0.5%, resulting in a relative resistance change of ∼0.01. Moreover, the reproducible operations of the GCF sensors were confirmed through the repeated vertical pressure and lateral strain. The sensor provided highly consistent responses for a repetitive operation of 6000 loading−unloading cycles corresponding to an applied vertical pressure of 40 kPa with a frequency of 1.7 Hz and a 10 ms measurement interval (Figure 2e). The sensor operated consistently in 1000 stretching−releasing cycles with a tensile strain of 20% (Figure 2f). For the bending strain, the sensitivity, (ΔR/R0)/κ, was estimated to be ∼0.015, where κ D


Research Article


Figure 4. Monitoring of various human activities. (a) Measurements of electrocardiogram (ECG) signals under dry and underwater conditions with bending up to 120° using the GCF attached on a wrist. (b) Detection of wrist-bending motion on both dry and wet skin surfaces. (c) Detection of radial artery pulse waves with the GCF sensor attached on a wrist. (d, e) Recognition of speech vibrations “I am good man” with the GCF sensor attached on the neck of the test person.

is the curvature, which confirms that this GCF sensor was flexible enough to be used in soft electronic devices (see Figure S9, Supporting Information). Moreover, the GCF sensors were highly responsive to the torsion forces that were applied by twisting the fixed ends of the sample in opposite directions. With a torsion sensitivity of (ΔR/R0)/θ, where θ is the torsion angle, the sensitivity was estimated to be 0.0019 (see Figure S10, Supporting Information). The enhancement of adhesion forces in dry and wet conditions owing to the OPs is confirmed in Figure 3. The GCFs are highly adherent to the human skin, which may be wet and rough (Figure 3a). The OPs,33 known to be a strong dry adhesive material on the skin, were patterned on the side of the GCF that touches the skin and made conformal contact with the human skin without additional chemicals. With the OPs formed on the fabric (Figure 3b), dry/wet normal adhesion strength is improved. In dry condition, the suction stress is added to the van der Waals force due to the OP, as shown in inset images (top and cross-sectional views of individual OP) of Figure 3b. The suction stress and capillary force are applied dominantly in wet conditions.33 As shown in Figure 3c, the GCFs with OPs exhibit higher normal adhesion strength than those without OPs for both dry and wet conditions. Importantly, the GCFs with OPs maintained their high adhesion characteristic with a weight of 200 g even in underwater conditions (Figure 3d). Such a characteristic has not been achieved for any reported strain and pressure sensors based on the fiber, textile, or fabric until now. This indicates that the GCFs with OPs can be applied directly to a waterresistant and skin-attachable wearable sensor and electrode system more efficiently. Although the OPs make the unique morphology of the fabric fade, such a dry adhesive pattern allows strong dry/wet adhesion on various surfaces including the human skin because of a suction effect induced by the OPs.29

The excellent monitoring capability of various human activities with the skin-adhesive and strain-sensitive properties is demonstrated in Figure 4. The ECG signals were first measured to demonstrate direct usability for a health-care monitoring electrode. It is well known that ECG signals provide meaningful information for biomedical diagnosis related to cardiovascular diseases with analysis of peak signals. Two identical GCF electrodes and a ground electrode were attached on both wrists and the left ankle of a volunteer. Figure S11 (Supporting Information) presents the ECG signals for ∼8 s in dry conditions for GCF electrodes (lower) and for commercial ECG electrodes (upper). The measured ECG signals with GCF electrodes appeared to be very similar to results with commercial ECG electrodes. More importantly, the GCF samples with OPs could measure ECG signals and human body motions in both dry and wet environments. The ECG signals and body motion signals were detected while the continuous bending motions of the wrist, which was attached with the samples without using any chemicals, as shown in Figure 4a,b. With bending up to 120° using the GCF attached on the wrist, ECG signals had the same output not only on a dry surface but also on a wet surface. In particular, the measurement of the ECG signals on the wet skin could be enabled by strong wet adhesion of our GCF sample by the OPs. The wrist-bending motion was also detected on both dry and wet surfaces. The electrical responses were detected with the increase of the bending angle at intervals of 15° up to 120° in dry and wet environments. The resistance change rate (ΔR/ R0 ∼ 0.9) in dry conditions shows that the bending (120°) induces a tensile strain of ∼18% (see Figure 2b). Moreover, the GCF sample could detect wrist bending even in wet conditions, a feature that was hardly achievable in previously reported fiber (or fabric)-based sensors. This is attributed to the high stretchability and strong dry/wet adhesion properties of the GCF samples. E


Research Article


was used for suspensions with rGO and P3HT (20:1 weight ratio). See the reference for the subsequent processes.28 The rGO and PDMS mixture was then coated on the rGO-coated fabric using a shallow dip-coating. The composite mixture was degassed using a vacuum desiccator for 2 h. The OP polymeric master with microhole patterns (diameter 100 μm, depth 100 μm, and spacing 100 μm in a hexagonal arrangement to maximize the filling rate of patterns) was prepared to form the OPs on the passivated rGO-coated fabric.28 The GCF with OPs was obtained by replicating a reversed architecture against the OP polymeric master on one side of the passivated rGOcoated fabric.

The application of the GCF samples as a health-care monitoring sensor was also investigated. Radial artery pulse waves were measured, and the acquired pulse train signals are plotted in Figure 4c and Figure S12 (Supporting Information). The GCF sensors were attached on a wrist or an index finger and positioned on the radial artery of a healthy 35-year-old man without any chemical adhesives. The result indicated that the sensor could continuously monitor pulse beats and waveforms in real time. This is attributed to the sensor’s high sensitivity to a subtle pressure range (

Water-Resistant and Skin-Adhesive Wearable Electronics Using

Recommend Documents