Sweat-permeable electronic skin with a pattern of eyes for body temperature monitoring

Abstract

Human-machine interface has been considered as a prominent technology for numerous smart applications due to their direct communication between humans and machines. In particular, wearable electronic skins with a free form factor have received a lot of attention due to their excellent adherence to rough and wrinkled surfaces such as human skin and internal organs. However, most of the e-skins reported to date have some disadvantages in terms of mechanical instability and accumulation of by-products at the interface between the human skin and the device. Here, we report a mechanically stable e-skin via a newly designed pattern named the “eyes.” The ingeniously designed pattern of the eyes allowed mechanical stress and strain to be dissipated more effectively than other previously reported patterns. E-skin permeability of by-product was experimentally confirmed through sweat removal tests, showing superior sweat permeability compared to conventional e-skins. Finally, the real-time monitoring of the body temperature was carried out using our resistive-type thermometer in the e-skin.

Introduction

With the beginning of advanced information age, Internet of Things (IoT) technologies have received much attention in recent decades, including human-machine interface (HMI), artificial intelligence (AI), and machine learning (ML) as a tool for info-communication between objects. These IoT techniques have been intensively investigated by many researchers due to their easy accessibility, rapid data processing and wide applicability [1,2,3,4,5,6]. In particular, HMI technology has been highlighted for use in numerous smart applications (e.g. voice recording, touch panels, and wearable electronic devices, etc.) thanks to its direct connectivity between man and machine [7,8,9,10,11,12].

Among various types of HMI devices, wearable electronic skins (e-skins) have been considered as an emerging platform for personal electronics with their free form factor, which can be attached to uneven and corrugated surfaces such as human skin and internal organs [13,14,15,16,17,18]. Although the previously reported work has demonstrated health monitoring sensors in e-skins (e.g., electrocardiogram (ECG) electrodes, photoplethysmogram (PPG) devices, and heart rate sensors), they still have some issues including mechanical instability and by-product accumulation at the interface between the human skin and the device [19,20,21]. In detail, the e-skin attached to the skin becomes constantly stressed by the bending and stretching through body movements in daily life, leading to device degradation and inaccurate sensing properties [22,23,24,25]. In addition, human by-products from the skin surface affect the attached e-skin, resulting in unwanted device delamination and skin troubles [26,27,28]. These problems need to be solved to sustain the lives of vulnerable people who require continuous monitoring of their vital signs. For example, patients with heat allergies suffer life-threatening symptoms (e.g. heat stroke, heat cramps, heat syncope) from sudden increases in body temperature during daily activities [29,30,31]. The elderly, infirm and infants also require constant physiological monitoring due to their lack of control over their body temperature [32,33,34].

Herein, we introduce a perforated e-skin with an eyes-patterned structure to monitor body temperature in real time. The eyes-pattern in the e-skin was theoretically modelled using finite element analysis (FEA) simulation to enhance mechanical properties and sweat permeability compared to other structures, such as auxetic kirigami and circular hole patterns. The resistive-type temperature sensor was fabricated at the center of a 10 μm-thick polyimide (PI) thin film to minimize stress localization. The developed e-skin maintained its mechanical and electrical performance under bending conditions with a radius of curvature of 500 μm. The sweat permeability of the developed device was analyzed by verifying the color change of litmus paper on the e-skin attached to the skin. Finally, we monitored the body temperature in real time using the resistive-type body temperature monitoring (BTM) patch, showing the coincidence of tendency with the conventional thermometer.

Result and discussion

Figure 1a shows the fabrication process and a three-dimensional schematic of the eyes-patterned e-skin. The detailed process is as follows: (i) 200 nm-thick aluminum (Al) and 5 μm-thick polyimide (PI) thin films were sequentially deposited on a rigid p-type Si substrate. (ii) Cr and Au thin films (thickness of 50 and 300 nm, respectively) were deposited and patterned by conventional complementary metal-oxide-semiconductor (CMOS) process to fabricate electrode and resistive temperature sensor. After covering the metal thin films with the PI passivation layer, a 200 nm-thick Cu etching mask with the desired patterns such as circular holes, kirigami, or eyes-patterns was fabricated on the PI surface, and then the whole layers were patterned by inductively coupled plasma-reactive ion etching (ICP-RIE) to provide the permeability for the e-skin. (iii) The ultrathin 10 μm-thick device was attached to a temporary substrate (thermal release tape; TRT), and the device was peeled off from the mother Si substrate through an electrochemical lift-off (ECL) process. The Al sacrificial layer on the mother wafer was removed by electrochemical reaction in 0.9 wt% NaCl solution at 3 V bias. (iv) After transferring the device to the TRT substrate, a polydimethylsiloxane (PDMS) adhesive layer was coated onto the back side of the e-skin. Finally, the temporary substrate was delaminated at 120 ℃ (See the Experiments section for a detailed description of the e-skin fabrication process). Figure 1b is an optical image of the fabricated eyes-patterned e-skin with body temperature monitoring (BTM) sensor electrically interconnected with Cu wires through Ag paste. The Cr-based resistive temperature sensor was successfully fabricated on the eyes-patterned e-skin as shown in Fig. 1c.

Fig. 1

a Fabrication process of eyes-patterned e-skin with BTM sensor. b Optical image of the wearable BTM sensor. c Magnified microscopic image of the temperature sensor in the developed e-skin

Because the skin-attached e-skin is constantly stressed by bending and stretching motions during daily activities, the e-skin need to provide mechanical stability to prevent the device from breaking or malfunctioning [35,36,37,38]. A novel eyes-pattern was developed to prevent the aforementioned issues in advance. Prior to the experiment, the mechanical stability of the eyes-pattern was theoretically analyzed by FEA simulation to compare with that of other patterns. To calculate the localized stresses and strains of the designed patterns in the e-skin, FEA calculations were performed using Ansys Workbench. Three typical geometric (TG) models, eyes, circular holes, and auxetic kirigami patterns were designed with the same scale in hole diameter, line width, and pattern distribution. The static structure calculation sequence was used to evaluate the von Mises stress using Eq. 1 below:

$$\sigma _{{vm}} = ~\sqrt {\frac{{\{ \left( {\sigma _{1} - ~\sigma _{2} } \right)^{2} + \left( {\sigma _{2} - ~\sigma _{3} } \right)^{2} + \left( {\sigma _{3} - ~\sigma _{1} } \right)^{2} }}{2}}$$

(1)

where ${\sigma }_{1}$, ${\sigma }_{2}$, and ${\sigma }_{3}$ are the stress components of each axis in three dimensions. The 45 μm displacement was applied to both ends of each geometry to observe the changes in localized stress and strain. The linear elastic property was used for the PI material (material density: 1.42 g/cm³, Young’s modulus: 2.5 GPa, and Poisson’s ratio: 0.34).

Figure 2a depicts the calculated stress distribution images of the e-skins with various patterns. Less than 30% of the applied strain conditions, all patterns in the 300 × 300  µm² scale (minimum line width: 12 μm, hole diameter: 75 μm) showed stress localization at the edge of the structure. However, as shown in Fig. 2b-i, the eyes-pattern exhibited a maximum localized stress of about 943.3 MPa, which was 48.7% and 60.8% lower than the auxetic kirigami (about 1839.9 MPa) and circular hole patterns (about 2408.8 MPa), respectively. Figure 2b-ii displays the localized strain of the patterns, showing that the eyes-pattern has the lowest localized strain of 43.11% compared to that of other structures (auxetic kirigami: 74.53%, circular hole: 101.41%), shown in Additional file 1: Table S1. These results indicate that the newly designed eyes-pattern has superior mechanical properties for stress relief and prevention of crack initiation and propagation compared to normal PI film (maximum strain: $<$ 20%) [39]. In order to experimentally confirm the mechanical stability of the patterns, the bending test was performed with various radii of curvature using an Arduino-based bending machine, as shown in Fig. 2c. Figure 2d depicts that the resistance changes of the eyes-patterned electrode were negligible due to the efficient stress dissipation along the entire structure, whereas the auxetic kirigami and circular hole-patterned e-skins were broken at the bending condition of 3 and 5 mm, respectively. Since the minimum bending radius of the human body is known to be about 3.8 mm [40, 41], it is noteworthy that the eyes-patterned e-skin can be considered as an optimal structure for excellent stress-resistant wearable e-skin during body movements.

Fig. 2

a Simulated stress mapping images of various e-skin structures. b (i) Comparison graph of the calculated maximum stress and (ii) strain in three different e-skins. c Optical images of the bending test (left) and various e-skin structures (right, scale bar: 500 μm). d Bending test results for each of the e-skin structures

In order to build practical wearable devices for real-time monitoring of vital signs, the accuracy of the sensing device without malfunctioning due to accumulation of skin by-products at the skin/e-skin interface is critical. As shown in Fig. 3a, our eyes-pattern in the e-skin has the periodic and porous structure with auxetic kirigami pattern, which was predicted to efficiently remove by-products at the skin-device interface due to similar size and distribution with human sweat pores [27]. To experimentally verify the by-product permeability of our e-skin, penetration tests of sweat, one of the representative skin by-products, were conducted through the eyes-patterned e-skin and a conventional e-skin for an hour. The amount of sweat penetration was determined by comparing the color changes of the litmus paper between our e-skin and the conventional e-skin. Figure 3b displays the experimental setup for the sweat penetration tests. The litmus papers on the conventional e-skin and our e-skin were attached to a 3 × 3 mm²-sized hole in the nitrile glove, which was used to accelerate the perspiration of the hand. Figure 3c shows the time-lapse images of the color changes of litmus papers attached to our e-skin and the conventional e-skin. Although the litmus paper on the conventional e-skin exhibited no color change, the litmus paper on our e-skin was wetted by excessive perspiration, showing color changes from yellow (pH 7) to green (pH 8). This result was attributed that our e-skin prevented sweat accumulation [42]. The area of color change was quantified using ImageJ, an image analysis tool. During a one hour-long perspiration test, the area percentage of color change in the litmus paper on our e-skin changed to 45.04%, which was remarkable compared to that on the conventional e-skin (Additional file 1: Fig. S1). These results demonstrate that our e-skin has notable sweat permeability and is suitable for daily activation without accumulation of skin by-products.

Fig. 3

a Sweat-permeability of eyes-patterned e-skin compared to conventional e-skin. b Optical image of the sweat-permeability test of the e-skins. The inset image shows a magnified microscopic image of the eyes-structured e-skin. c Color-changed images of litmus paper to monitor sweat permeability of e-skins (scale bar: 500 μm)

Finally, our e-skin with excellent mechanical stability and sweat permeability was applied to real-time body temperature monitoring via a resistive-type BTM sensor. The reliability of our BTM sensor was verified by comparison with a conventional infrared (IR) thermometer (Fig. 4a). As the temperature increased from 20 ℃ to 90 ℃, the BTM sensor showed a linear increase in resistance from 204 Ω to 682 Ω, showing the same trend as a conventional IR thermometer (average resistance increase ratio is 6.82 Ω K^− 1). To investigate the suitability of the developed wearable sensor for real-time body temperature monitoring, the skin adherence test was performed for periodic exercise and rest cycles. Figure 4b is a photograph of a 1 $\times$ 1 cm²-sized wearable BTM sensor attached to the skin surface of a human forearm. Despite periodic push-up exercises, the skin-attached e-skin remained stable in position with no device delamination (Fig. 4c). During the repeated exercise, the wearable BTM sensor clearly exhibited the resistance difference between the exercise (325.6 Ω) and rest (281.5 Ω) conditions, showing the same trend with the conventional IR thermometer, as shown in Fig. 4d. The developed BTM sensor had a similar body temperature sensing capability at the skin surface, while having a simpler circuit structure than conventional thermometers. Based on these results, our wearable BTM sensor with sweat permeability may be a potential candidate to replace conventional body temperature monitoring systems.

Fig. 4

a Temperature sensing performance of the BTM sensor. Optical images of b the BTM sensor on the forearm skin and c exercising with the BTM sensor. d Skin temperature was monitored by the BTM sensor during exercise compared to a conventional thermometer

Conclusion

In summary, we have developed a novel e-skin with a unique eyes-pattern that provides superior mechanical stability and by-products permeability. Our eyes-pattern, which was sophisticatedly designed by FEA simulation, exhibited the lowest localized stress of 943.3 MPa and strain of 43.1% compared to the previously reported patterns including auxetic kirigami and circular holes. While other patterned e-skins showed electrical malfunctions due to electrode breakdown under the bending radius changes from flat to 500 μm, our e-skin successfully maintained its electrical properties without any electrode disconnection. In order to confirm the permeability of the device to by-products, simple sweat elimination tests were performed on the e-skin using litmus paper. During a one-hour long test, the litmus paper color on our e-skin changed about 45.04% by area due to sweat penetration, while that on conventional e-skins remained unchanged. Real-time temperature monitoring test using our eye-patterned e-skin was successfully performed during harsh body movements of periodic push-up cycles. The resistive-type BTM sensor was formed on our e-skin, which has a linear resistance ratio of about 6.82 Ω K^− 1. During the periodic exercise and rest cycles, the wearable BTM sensor was stably attached to the human skin surface and monitored the body temperature with the same trend as a conventional IR thermometer. We believe that with our eyes-patterned e-skin, we may have proposed a new approach for reliable real-time healthcare applications.

Experiments

Fabrication of the patterned e-skin

The Al sacrificial layer was deposited on a 4-inch Si substrate by thermal evaporation (VEV-503, VTS Corporation). Next, polyimide solution (PI, Sigma-Aldrich) was spin-coated onto the Si surface at 3000 rpm and baked sequentially at 110 ℃ (3 min), 130 ℃ (3 min), and 250 ℃ (1 h) to cure the film by evaporating the solvent. An Au electrode layer (300 nm) was deposited on the bottom PI by e-beam evaporation and patterned by conventional CMOS processes. A resistive metal layer (Cr, 20 nm) of 100 × 100 µm² was formed by interconnecting the Au electrodes to fabricate the BTM sensor. After coating the top PI layer, e-beam evaporation was used to create a Cu mask layer (200 nm). The ICP-RIE process was carried out under optimized conditions (50 sccm O₂ flow, 20 mTorr working pressure, 300 W ICP and 150 W bias). After the ECL process, the PDMS adhesion layer was applied to the backside of the delaminated e-skin and then cured at 70 ℃ for one hour.

Evaluation of mechanical properties

The Cu-wire was electrically connected to the metal pads of the e-skin with silver epoxy (MG Chemicals, 8331 S). To verify electrical performance under various bending radii from flat to 0.5 mm, a conventional linear stage motor (LSM-NK235603, Motorbank) and source meter (Keithley 2450, Tektronix) were used at 1 V bias. The bending curvature radius was accurately controlled from flat to 0.5 mm through the programmed codes in the bending motion system.

Evaluation of the BTM performance

The resistive-type BTM sensor was integrated into the e-skin. The BTM sensor attached to the forearm measured the change in resistance of the sensor during exercise and reset in real time. For continuous perspiration, we fulfilled the push-up 30 times at each cycle and had 5 min resting time.