Article pubs.acs.org/acssensors
A Wearable and Highly Sensitive Graphene Strain Sensor for Precise Home-Based Pulse Wave Monitoring Tingting Yang,†,‡ Xin Jiang,†,‡ Yujia Zhong,†,‡ Xuanliang Zhao,†,‡ Shuyuan Lin,†,‡ Jing Li,† Xinming Li,§ Jianlong Xu,∥ Zhihong Li,*,‡,⊥ and Hongwei Zhu*,†,‡ †
State Key Laboratory of New Ceramics and Fine Processing, School of Materials Science and Engineering and ‡Center for Nano and Micro Mechanics, Tsinghua University, Beijing 100084, China § Department of Electronic Engineering, The Chinese University of Hong Kong, Hong Kong SAR, China ∥ Institute of Functional Nano and Soft Materials (FUNSOM), Jiangsu Key Laboratory for Carbon-based Functional Materials and Devices, Soochow University, Suzhou 215123, Jiangsu, China ⊥ National Key Laboratory of Science and Technology on Micro/Nano Fabrication, Institute of Microelectronics, Peking University, Beijing 100871, China S Supporting Information *
ABSTRACT: Profuse medical information about cardiovascular properties can be gathered from pulse waveforms. Therefore, it is desirable to design a smart pulse monitoring device to achieve noninvasive and real-time acquisition of cardiovascular parameters. The majority of current pulse sensors are usually bulky or insufficient in sensitivity. In this work, a graphene-based skin-like sensor is explored for pulse wave sensing with features of easy use and wearing comfort. Moreover, the adjustment of the substrate stiffness and interfacial bonding accomplish the optimal balance between sensor linearity and signal sensitivity, as well as measurement of the beat-to-beat radial arterial pulse. Compared with the existing bulky and nonportable clinical instruments, this highly sensitive and soft sensing patch not only provides primary sensor interface to human skin, but also can objectively and accurately detect the subtle pulse signal variations in a real-time fashion, such as pulse waveforms with different ages, pre- and post-exercise, thus presenting a promising solution to home-based pulse monitoring. KEYWORDS: strain sensor, pulse diagnosis, wearable, highly sensitive, graphene
P
Currently, there are several types of sensors that are used to acquire the pulse signal such as piezoelectricity sensor, infrared sensor, acoustic sensor, liquid sensor, photoelectric sensor, and image sensor.11−15 In traditional Chinese pulse diagnosis, physicians pick up the human pulse information by means of fingertips palpating patients’ pulses shown in the superficial arteries. Hence, a pressure or strain sensor may be the best option to imitate the physician’s tactile sensation of the pulse in the practice of traditional Chinese medicine.16−23 Since the pulse wave belongs to weak human motions, it is difficult for sensors to differentiate all features precisely in a characteristic pulse waveform shape. The majority of current pulse sensors
ulse waves, consisting of the systolic pressure wave and vascular elasticity damped wave, have a strong correlation with cardiovascular events and reveal significant pathological information.1−4 In traditional Chinese medicine, the pulse diagnosis has achieved great success at deducing the locations and degree of pathological changes during the long period of clinical practice.5,6 In Western medical examination, the rate, rhythm, wave velocity, and volume of the pulse wave also indicate the status of the heart and blood vessels.7,8 Different from an electrocardiogram (ECG) that reflects the bioelectrical information in the body, the pulse signals originate from the pressure fluctuation of the artery and typically possess a waveform shape with three clearly distinguishable peaks.9 Through appropriate signal processing and pattern recognition methods, the pulse signal can even reveal some diagnostic information that ECG cannot.10 © 2017 American Chemical Society
Received: April 6, 2017 Accepted: July 5, 2017 Published: July 5, 2017 967
DOI: 10.1021/acssensors.7b00230 ACS Sens. 2017, 2, 967−974
Article
ACS Sensors
Figure 1. System design of the wearable pulse sensor and home-based health monitoring system. (A) Schematic illustration of the pulse sensor placed on a subject’s wrist. (B) Schematic structure of the wearable platform for home-based health monitoring. (C) Photographs of the continuous pulse measurement.
are usually bulky or insufficient in sensitivity. Thus, equipment (belt or cuff) attached to the pulse sensor on the wrist is required to avoid information loss.21,24 However, the pressure of the belt or cuff would cause discomfort possibly causing the blood in the vein to flow backward.10 Therefore, it is desirable to design new materials or devices to enhance the sensor sensitivity and wearability for continuous, real-time, and elaborate pulse monitoring at home. Graphene is a promising active material for the development of advanced sensing devices.25−31 It has been reported that a graphene-based sensor can detect the pulse rate effectively.32,33 However, due to low sensitivity or inappropriate device structures, accurately recorded and detailed analysis of the pulse waveform shape is deficient to the best of our knowledge. In this study, we describe a wearable and highly sensitive graphene strain sensor for home-based pulse wave measurement. Graphene woven fabrics, as previously reported, possess an extremely high gauge factor of about 500 under 2% strain, suggesting that it is a promising sensing material for the detection of subtle deformation.34 However, its approximately exponential rise in resistance under strain may produce negative consequences such as waveform distortion. To overcome this drawback, we investigated the effect of the substrate stiffness on the sensor sensitivity and reached an acceptable compromise between sensor linearity and sensitivity. After the optimization, the graphene sensor was attached to a volunteer’s wrist, just above the radial artery, using a 3 M adhesive tape to demonstrate beat-to-beat radial arterial pulse measurement. Because of the lightweight and flexible properties, the graphene sensor could follow skin deformation easily without causing any
feelings of irritation. Noninvasive pulse waveform shape recordings in four healthy individuals were conducted to illustrate the age-related changes in wave reflection characteristics. Three of the most commonly used parameters for pulse diagnosis including the time delay between the percussion wave and tidal wave (ΔtDVP), the radial augmentation index (AIr), and the auxiliary blood pressure index (k) were extracted. The result revealed a regular dependence on age despite personally different features of the subject. Pulse waveform rate, magnitude, and shape for one female volunteer before and after exercise also showed differentiable results. Moreover, a row of three sensors was located on the inch openings including the Cun (distal), Guan (middle), and Chi (proximal) positions for pulse taking simultaneously. The sensors at these three positions can differentiate all the characteristic peaks in the diastolic tail of the pulse wave shape, and hence could potentially be of use in multiposition pulse diagnosis in the future.
■
RESULTS AND DISCUSSION Sensor Fabrication and Home-Based Health Monitoring System. Figure 1A shows the schematic illustration of the sensor structure. The key component graphene woven fabrics (GWFs) was synthesized by chemical vapor deposition (CVD) on copper meshes.35 The as-fabricated graphene has a crisscross structure as shown in Figure S1. Then, a thin film of polydimethylsiloxane (PDMS) with 100 μm thickness, a biocompatible and highly elastic polymer, was used to collect the GWFs. Thereafter, a silver adhesive was coated on the two ends of GWFs to make an electrical connection. The as968
DOI: 10.1021/acssensors.7b00230 ACS Sens. 2017, 2, 967−974
Article
ACS Sensors
Figure 2. Tuning of the sensor sensitivity versus linearity and their effects on pulse monitoring quality. (A) Electromechanical response of GWFs on different PDMS substrate of varied stiffness under stretch. (B−E) Raw pulse signal and enlarged view of the signal for illustrating the pulse waveform (inset) using pulse sensor with different combination of sensitivity and linearity: (B) GF = 5, R2 = 0.991; (C) GF = 20, R2 = 0.971; (D) GF = 50, R2 = 0.926; (E) GF = 200, R2 = 0.917. (F) Epidermal pulse amplitude and linearity versus sensor sensitivity.
fabricated graphene strain sensor exhibited an extremely high gauge factor of 500 (0−2%), 103 (2−6%), and 106 (>8%) because of its unique crisscross morphology and crack propagation mode.34,36 Then, the whole strain sensor was attached to the radial artery location of the human wrist using a 3 M adhesive tape (50 μm thickness, good elasticity) to collect the pulse waveform objectively. Figure 1B demonstrates a schematic of the smart sensor system for home health care. The resistance data of the sensor vary with the fluctuation of the artery pressure and get processed and transmitted using a custom silicon-based circuit system. The silicon-based circuit system includes an analog front end which filters and amplifies the pulse signal and a USB or Bluetooth module which further digitizes and transmits these data to a host computer or cellphone. Finally, the collected data are displayed on the liquid crystal display screen of the computer or cellphone through a custom user interface. The complete system is powered by USB cable or external 5 V power supply. Figure 1C presents photographs of the continuous pulse measurement using a home-based health monitoring system with virtues of light weight, handy volume, and friendly user-interface. Their corresponding movies (Movies S1, S2) are provided in the Supporting Information. Adjustment of the Substrate Stiffness to Achieve an Optimal Trade-Off between Sensitivity and Linearity for Pulse Monitoring. For pulse sensor, linearity is another critical parameter because good linearity means that each change in the pressure fluctuation of the artery leads to the same change in the electrical output of the sensor. Such undistorted output of a sensing device facilities the calibration process and reliable extraction of pulse waveform characteristic parameters. In this work, we explore the effect of substrate on the linearity of the GWFs based sensor. It was found that the transition from nonlinear mode to approximately linear mode could be predicted by the stiffness ratio between the substrate
and the GWFs. As shown in Figure 2A, when tuning the ratios of base to cross-linker from 20:1 to 8:1, namely, increasing Young’s modulus of the substrate PDMS from hundreds of kPa to a few MPa, the corresponding sensitivity decreased while the linearity was improved with the linear fitting parameter of R2 (adjusted R-square) ranging from 0.911 to 0.981. The reason for the diminished sensitivity is ascribed to the produced buckle-delamination in GWFs (Figure S2A), in which the stiff substrate is stretched initially and then released. In contrast, less buckle-delamination (more like wrinkling) was observed for graphene on soft PDMS (Figure S2B). The formation of buckle-delamination in GWFs when the stiff substrate is used could be explained as follows. For soft substrate, the substrate is relatively compliant; thus it is possible to form a wrinkling pattern of graphene which requires coherent deformation of the substrate. On the contrary, for the stiff substrate, from an energy point of view, it is unfavorable to form a wrinkling pattern of graphene due to the high critical stresses for wrinkling. Instead, the interfacial defects may lead to partial delamination of graphene which in turn drives the growth of the buckle-delamination pattern through the interfacial fracture. A theoretical and an experimental understanding of the buckling mode selection for elastic thin films on the elastic substrate have been reported.37,38 Such buckle-delamination successfully converts graphene stretch into bending for the subsequent tensile tests, which transfers the crack dominant transduction principle to a geometry dominant one, thus leading to improved linearity. In other words, there is a compromise between high sensitivity and high linearity for the piezoresistive-type sensor. Notably, our sensor exhibited an optimal balance between these two criteria. For example, when fixing the linearity to be R2 ≈ 0.98, the designed graphene sensor exhibited a higher sensitivity (GF ≈ 12) compared with a strain sensor based on silver nanowire-elastomer nanocomposite (GF ≈ 2). 969
DOI: 10.1021/acssensors.7b00230 ACS Sens. 2017, 2, 967−974
Article
ACS Sensors
Figure 3. Epidermal pulse monitoring with different ages. (A) Illustration of the pulse generation and its waveform shape features. (B) Three healthrelated parameters for subjects of different ages. (C−E) Original pulse and its frequency spectra distribution for male volunteers of (C) age 10 yrs; (D) age 23 yrs; (E) age 62 yrs, the black arrows indicate the peak position of a tidal wave.
left ventricular. The descending limb contains two characteristic waves caused by blood flow reflection, namely, tidal wave from the hand region and a later-arriving dicrotic wave from the lower body.9 Since pulse formation originates from the cardiovascular activity, some cardiovascular factors, such as cardiac contractility and ejection fraction, peripheral resistance and artery compliance, have a notable influence on the pulse. Through qualitative analysis of the pulse wave, some key parameters can be easily derived for the evaluation of health conditions.39,40 Herein, three of the most commonly used parameters for quantitative pulse analysis are presented,9,40,41 including the time delay between the percussion wave and tidal wave (ΔtDVP), the radial augmentation index (AIr = P2/P1), and the auxiliary blood pressure index (k), defined by the following equation:
To achieve a highly reliable epidermal pulse monitoring with minimum feature loss, we introduced two parameters, namely, the sensor sensitivity and linearity, to evaluate the pulse monitoring quality. As shown in Figure 2B, when the sensitivity was low (GF ≈ 5), the pulse rate was reduced, but the waveform was unstable, and its shape pattern was hard to be recognized. When the GF reached 20 or more as shown in Figure 2C−E, all the three pulse shapes displayed stable and clearly distinguishable peaks. These results indicate that high sensitivity is a prerequisite for the reliable pulse sensing. As shown in Figure 2F, the pulse amplitude increased with increasing sensor sensitivity, but the linearity was degraded. To obtain an optimal pattern recognition of pulse shape along with minimum distortion, we fixed the GF of the sensors used in the following experiments at around 20 (R2 ≈ 0.97). A comparison between the GWFs-based pulse sensor and other reported results has been shown in Table S1. Qualitative Analysis of Pulse Wave Associated with the Changes of Age. The radial artery is the most common location for pulse diagnosis. A normal radial artery pulse is composed of an ascending limb and a descending limb as illustrated in Figure 3A.10 As a part of the ascending limb, the percussion wave is caused by the incoming blood flow from the
k=
Pm − Pd Ps − Pd
where Pm is average epidermal pulse amplitude. Clinical studies indicate that age is an important determinant of the elastic properties of the arterial wall, and a series of natural changes in arterial pulse waves emerge along with aging.42 Since the elastic arteries become stiffer with age, the 970
DOI: 10.1021/acssensors.7b00230 ACS Sens. 2017, 2, 967−974
Article
ACS Sensors
Figure 4. Epidermal pulse monitoring pre- and post-exercise. (A) Schematic of a female volunteer taking squat exercise. (B) Recorded pulse data at still and exercise state. (C) Enlarged view of the dashed box region to show the waveform shape features pre- and post-exercise.
within the normal range (0−10 Hz) for a healthy person.43 The above pulse analysis may still be too crude to reflect the health conditions of the cardiovascular system because it ignores the individual difference, but the capability of our sensor to distinguish the tiny differences among various pulse waveforms is promising. Test of Anti-Interference Ability. In practical usage, the anti-interference ability of the pulse is vitally important, i.e., whether the signal will be affected by human motions, such as swinging arms, flexing wrist, and finger touch. At the beginning of the anti-interference ability test of the pulse sensors, the volunteer held the seated posture motionless as mentioned earlier. Then she flexed her wrist for a few seconds and finally returned it to the original position. The corresponding electrical signal of the sensor was simultaneously recorded and shown in Figure S3. Before the motion interference, the pulse waveform variation collected from the sensor is stable and reproducible. When flexing the wrist, the skin covering the radial artery location is elongated, thus causing the increased resistance of the sensor. Since the wrist flex motion brings larger skin deformation than the inherent intensity of pulsation, the pulse signal is almost lost over this period of time. After motion interference was removed, the pulse cycles recovered with welldefined characteristic peaks. Therefore, it is suggested to select the stable-term signal to obtain the waveform parameters by the appropriate software algorithm in realistic applications. Pulse Wave Monitoring Pre- and Post-Exercise. The fabricated sensor could also collect the subtle variation of the pulse waveform pre- and post-exercise. A healthy female volunteer of 27 yrs was recruited for this trial. She would take squat exercises with a speed of 50 times per minute as illustrated in Figure 4a. Once the exercise was finished, the pulse wave was collected immediately using the sensor. The signals of the pulse wave at three body states of resting (black), 2 min of exercise (red), and 4 min of exercise (blue) are shown in Figure 4B. With the increase of exercise intensity, higher pulse rate and larger pulse amplitude were observed. The pulse rate in the resting state and two exercise states were 85, 108, and 129 per minute, respectively. Moreover, the primary waveform of the signal of exercise state (2 min) showed some changes compared with that of the resting state, as shown in Figure 4C. For example, the tidal wave became very close to the percussion wave and tended to be blurry. Simultaneously, the peak of the dicrotic wave gradually became unclear. After an extended period of exercise (4 min), the waveform (blue) became unstable and irregular. These results indicate high-intensity exercise would cause obvious
velocity of the pulse wave increases. Meanwhile, the reflected wave from the peripheral reflecting sites in the lower body returns earlier. These histologic changes lead to increased wave reflection amplitude and speed, thus causing a decrease in ΔtDVP and an increase in AIr.9 These changes in amplitude and timing of pulse wave also affect the blood pressure index k.40 To test the feasibility of the designed sensors in the qualitative analysis of pulse wave associated with the changes of age, we recruited five healthy male volunteers with different ages of 10, 23, 28, 43, and 62 in the trials. All the volunteers were healthy without physiological obesity or disabilities, and they did not take any medicines or vigorous exercise within 24 h before the trial. For each test, each subject kept the seated posture with arms raised vertically with the body for 1 min before the measurement. To estimate the three parameters ΔtDVP, AIr and k, the measured periodic pulse waves in 1 min was averaged to get one representative pulse contour, from which the derived AIr (0.62−0.95) and ΔtDVP (0.09−0.15 s) exhibited positive and negative correlations with age, respectively, as shown in Figure 3B. The blood pressure index k for the volunteers of 10, 23, 28, and 43 years were very close, ranging from 0.343 to 0.353. However, the oldest volunteer of 62 yrs had a significantly higher k value of 0.390. The k value provides a convenient assessing index for individual health condition that could be characterized as “low resistance type” (k < 0.35), “medium resistance type” (0.35< k < 0.4), “high resistance type” (0.4< k < 0.45), and “ultra high resistance type” (0.45< k < 0.5).40 According to the relatively high value of k, it could be inferred that the oldest volunteer has a bigger risk of atherosclerosis than the other four. A group of pulse waves and its frequency distribution of 10 (Figure 3C), 23 (Figure 3D), and 62 (Figure 3E) years old are shown, respectively. These waveforms show a coherent change trend as the previous study reported.9 As an objective and critical indicator reflecting vascular elasticity and vascular resistance, the tidal wave gradually migrated up with increasing age and approached the main wave for the elderly (Figure 3E), leading to a larger angle of the percussion peak and shorter ΔtDVP. Meanwhile, the position of the dicrotic wave peak and notch gradually became mixed and unclear. The frequency spectrum distribution of different age was obtained through fast Fourier transformation (FFT) after normalizing the magnitude of the collected pulse waveforms. According to the position of the first main peak in FFT analysis, the pulse rates were calculated as 83, 77, and 66 per minute for the 10, 23, and 62 year old volunteers, respectively. All the distributions of resonance wave components of the pulse wave were also 971
DOI: 10.1021/acssensors.7b00230 ACS Sens. 2017, 2, 967−974
Article
ACS Sensors
Figure 5. Multiposition pulse reading. (A) Location of the inch opening. (B) Pulse signals of the sensors located at the Cun (distal), Guan (middle), and Chi (proximal) positions.
evaluation of cardiovascular properties, in which the majority of currently used sensors in arterial tonometry are incompetent. Multiposition pulse reading was also explored to provide more medical information. In light of the various advantages of this graphene sensor, such as the lightweight, simplicity of the device fabrication, optimal balance between sensitivity and linearity, and easy-to-follow human skin deformation without irritation, it is a promising candidate for noninvasive and realtime health monitoring at home.
variation of cardiovascular parameters within a short period of time, completely consistent with the realistic physiological behaviors in the previous clinical study. Multiposition Pulse Reading. There are different locations for pulse taking particularly where superficial arteries can be felt, such as the radial artery, brachial artery, carotid artery, and dorsal artery of the foot. Through putting several pulse sensors in these different positions to read pulses simultaneously, more information about the physical condition can be obtained. According to The Classic of Dif f icult Issues (Nan Jing), inch opening constitutes the meeting point of the vessels and contains information on the health and disease of the entire body. Hence traditional Chinese medicine prefers inch-opening palpating.10 Inspired by traditional Chinese medicine, we located a row of three sensors on the inch opening including the Cun (distal), Guan (middle), and Chi (proximal) positions for pulse taking simultaneously. These three positions are shown in Figure 5A. Their corresponding pulse signals were recorded and illustrated in Figure 5B. The pulse rates collected at these three positions exhibited almost the same value of about 82 per minute. Among them, the pulse amplitude at the Guan position was the largest while that at the Chi position was the smallest. In addition, the sensors at these three positions can differentiate all the characteristic peaks in the diastolic tail of the pulse wave shape, and hence could potentially be used for the diagnosis of multiposition in the future.
■
EXPERIMENTAL SECTION
Material and Pulse Sensor Fabrication. GWFs were synthesized by CVD using copper woven meshes (#100, wires of 60 μm in diameter) as the growing template as reported previously.35,44 After 20 min growth in a thermal furnace at 1000 °C and ambient pressure under a gas flow of Ar/H2/CH4 (200/10/30 mL/min), the mesh was rapidly cooled down to room temperature. The as-fabricated GWFs covered the surface of copper mesh and inherited its crisscross configuration. Then copper mesh was removed by an aqueous solution of 0.5 M FeCl3/HCl and GWFs was cleaned using deionized water. The fabrication of graphene-based pulse sensor is described as follows: First, a mixture of PDMS (Sylgard 184, Dow Corning, the weight ratio of base to cross-linker could be tuned) was stirred for at least 20 min and degassed under vacuum for another 20 min to remove bubbles. Then the PDMS mixture was spin-coated onto hydrophobically treated Si wafer and got solidified at 80 °C for 3 h to get a thin film of ∼100 μm. After peeling off from the Si wafer, the PDMS was used as the substrate to collect the GWFs (6 mm × 5 mm). Thereafter silver wires were connected using silver paste to the two ends of GWFs to make an electrical connection. Electromechanical Response Measurement of the Strain Sensor. The tensile test was carried out by a stretching machine (Instron 5943) with a strain rate of 0.1%/s. Meanwhile, the electrical response of the sensor was recorded by a digital SourceMeter (Keithley 2602) with 1 V DC bias. Epidermal Pulse Signal Measurement. The whole pulse sensor was attached to the location of pulse taking using a 3 M adhesive tape (50 μm thickness, good elasticity) to objectively collect the pulse waveform. Then the sensor resistance variation was recorded by a digital source meter (Keithley 4200-SCS) with a sampling frequency of 100 Hz. As for the home-based health monitoring system as shown in the SI movies, the sensor was connected to a custom-made siliconbased circuit to get a resistance−voltage conversion. Thereafter the built-in USB or Bluetooth module would further transmit the processed data to a host computer or cellphone. A custom user interface in the terminal was used to display the pulse waveform finally.
■
CONCLUSION We have demonstrated a graphene-based skin-like sensor for noninvasive and real-time pulse wave sensing. The use of graphene woven fabrics as the sensing element provides high sensitivity at little strain which is a prerequisite for reliable pulse diagnosis. Furthermore, we investigated the effect of the substrate stiffness on the electromechanical response of graphene and established a relationship between the device linearity and sensitivity. Such substrate effect can be applied to other piezoresistive devices for material design and performance optimization. Moreover, the adjustment of the substrate stiffness accomplishes an optimal balance between acceptable linearity and high sensitivity, which further realizes the beat-tobeat radial arterial pulse measurements for people of different ages and at pre- and post-exercise. According to these results, the fabricated sensor can reveal tiny feature variation in the pulse wave and easily derive physiological parameters for the 972
DOI: 10.1021/acssensors.7b00230 ACS Sens. 2017, 2, 967−974
Article
ACS Sensors
■
sensors using oriented aluminum nitride thin films prepared on polyethylene terephthalate films. J. Appl. Phys. 2006, 100 (11), 114318. (12) Shao, X.; Dai, X.; Chen, Z.; He, X. Real-time 3D digital image correlation method and its application in human pulse monitoring. Appl. Opt. 2016, 55 (4), 696−704. (13) Chang, R.; Chen, Y.; Jwo, K.; Chen, D.; Hsieh, Y. Application of automatized 3D moiré monitoring system in pulse measurement. Opt. Express 2015, 23 (11), 14044−14056. (14) Li, Z.; Wang, Z. L. Air/Liquid-Pressure and Heartbeat-Driven Flexible Fiber Nanogenerators as a Micro/Nano-Power Source or Diagnostic Sensor. Adv. Mater. 2011, 23 (1), 84−89. (15) Park, J.; Kim, M.; Lee, Y.; Lee, H. S.; Ko, H. Fingertip skin− inspired microstructured ferroelectric skins discriminate static/ dynamic pressure and temperature stimuli. Sci. adv. 2015, 1 (9), e1500661. (16) Schwartz, G.; Tee, B. C.-K.; Mei, J.; Appleton, A. L.; Kim, D. H.; Wang, H.; Bao, Z. Flexible polymer transistors with high pressure sensitivity for application in electronic skin and health monitoring. Nat. Commun. 2013, 4, 1859. (17) Zang, Y.; Zhang, F.; Huang, D.; Gao, X.; Di, C.; Zhu, D. Flexible suspended gate organic thin-film transistors for ultra-sensitive pressure detection. Nat. Commun. 2015, 6, 6269. (18) Li, X.; Yang, T.; Yang, Y.; Zhu, J.; Li, L.; Alam, F. E.; Li, X.; Wang, K.; Cheng, H.; Lin, C. T.; et al. Large-Area Ultrathin Graphene Films by Single-Step Marangoni Self-Assembly for Highly Sensitive Strain Sensing Application. Adv. Funct. Mater. 2016, 26 (9), 1322− 1329. (19) Yang, T.; Li, X.; Jiang, X.; Lin, S.; Lao, J.; Shi, J.; Zhen, Z.; Li, Z.; Zhu, H. Structural engineering of gold thin films with channel cracks for ultrasensitive strain sensing. Mater. Horiz. 2016, 3 (3), 248−255. (20) Park, H.; Jeong, Y. R.; Yun, J.; Hong, S. Y.; Jin, S.; Lee, S.-J.; Zi, G.; Ha, J. S. Stretchable array of highly sensitive pressure sensors consisting of polyaniline nanofibers and Au-coated polydimethylsiloxane micropillars. ACS Nano 2015, 9 (10), 9974−9985. (21) Wang, C.; Li, X.; Gao, E.; Jian, M.; Xia, K.; Wang, Q.; Xu, Z.; Ren, T.; Zhang, Y. Carbonized Silk Fabric for Ultrastretchable, Highly Sensitive, and Wearable Strain Sensors. Adv. Mater. 2016, 28 (31), 6640−6648. (22) Wang, Y.; Wang, L.; Yang, T.; Li, X.; Zang, X.; Zhu, M.; Wang, K.; Wu, D.; Zhu, H. Wearable and highly sensitive graphene strain sensors for human motion monitoring. Adv. Funct. Mater. 2014, 24 (29), 4666−4670. (23) Shi, J.; Li, X.; Cheng, H.; Liu, Z.; Zhao, L.; Yang, T.; Dai, Z.; Cheng, Z.; Shi, E.; Yang, L.; et al. Graphene reinforced carbon nanotube networks for wearable strain sensors. Adv. Funct. Mater. 2016, 26 (13), 2078−2084. (24) Rajzer, M. W.; Wojciechowska, W.; Klocek, M.; Palka, I.; Brzozowska-Kiszka, M.; Kawecka-Jaszcz, K. Comparison of aortic pulse wave velocity measured by three techniques: Complior, SphygmoCor and Arteriograph. J. Hypertens. 2008, 26 (10), 2001− 2007. (25) Huang, D.; Yang, Z.; Li, X.; Zhang, L.; Hu, J.; Su, Y.; Hu, N.; Yin, G.; He, D.; Zhang, Y. Three-dimensional conductive networks based on stacked SiO 2@ graphene frameworks for enhanced gas sensing. Nanoscale 2017, 9 (1), 109−118. (26) Hu, N.; Yang, Z.; Wang, Y.; Zhang, L.; Wang, Y.; Huang, X.; Wei, H.; Wei, L.; Zhang, Y. Ultrafast and sensitive room temperature NH3 gas sensors based on chemically reduced graphene oxide. Nanotechnology 2014, 25 (2), 025502. (27) Wang, T.; Huang, D.; Yang, Z.; Xu, S.; He, G.; Li, X.; Hu, N.; Yin, G.; He, D.; Zhang, L. A review on graphene-based gas/vapor sensors with unique properties and potential applications. Nano-Micro Lett. 2016, 8 (2), 95−119. (28) Chen, Z.; Wang, Z.; Li, X.; Lin, Y.; Luo, N.; Long, M.; Zhao, N.; Xu, J.-B. Flexible Piezoelectric-Induced Pressure Sensors for Static Measurements Based on Nanowires/Graphene Heterostructures. ACS Nano 2017, 11 (5), 4507−4513. (29) Li, X.; Tao, L.; Chen, Z.; Fang, H.; Li, X.; Wang, X.; Xu, J.-B.; Zhu, H. Graphene and related two-dimensional materials: Structure-
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acssensors.7b00230. SEM images; anti-interference ability test; comparison between the GWFs based pulse sensor and other reported results (PDF) Home-based health monitoring system (laptop) (AVI) Home-based health monitoring system (cellphone) (AVI)
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail:
[email protected]. *E-mail:
[email protected]. ORCID
Xinming Li: 0000-0002-7844-8417 Hongwei Zhu: 0000-0001-6484-3371 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (51672150, 51372133), the National Key Research and Development of China (2016YFA0200802).
■
REFERENCES
(1) Wilkinson, I. B.; Hall, I. R.; MacCallum, H.; Mackenzie, I. S.; McEniery, C. M.; van der Arend, B. J.; Shu, Y.-E.; MacKay, L. S.; Webb, D. J.; Cockcroft, J. R. Pulse-wave analysis. Arterioscler., Thromb., Vasc. Biol. 2002, 22 (1), 147−152. (2) Van Bortel, L. M.; Laurent, S.; Boutouyrie, P.; Chowienczyk, P.; Cruickshank, J.; De Backer, T.; Filipovsky, J.; Huybrechts, S.; MattaceRaso, F. U.; Protogerou, A. D.; et al. Expert consensus document on the measurement of aortic stiffness in daily practice using carotidfemoral pulse wave velocity. J. Hypertens. 2012, 30 (3), 445−448. (3) Wang, X.; Gu, Y.; Xiong, Z.; Cui, Z.; Zhang, T. Silk-molded flexible, ultrasensitive, and highly stable electronic skin for monitoring human physiological signals. Adv. Mater. 2014, 26 (9), 1336−1342. (4) Boutry, C. M.; Nguyen, A.; Lawal, Q. O.; Chortos, A.; RondeauGagné, S.; Bao, Z. A sensitive and biodegradable pressure sensor array for cardiovascular monitoring. Adv. Mater. 2015, 27 (43), 6954−6961. (5) Shu, J.; Sun, Y. Developing classification indices for Chinese pulse diagnosis. Complement. Ther. Med. 2007, 15 (3), 190−198. (6) Lu, A.; Jiang, M.; Zhang, C.; Chan, K. An integrative approach of linking traditional Chinese medicine pattern classification and biomedicine diagnosis. J. Ethnopharmacol. 2012, 141 (2), 549−556. (7) Yamashina, A.; Tomiyama, H.; Takeda, K.; Tsuda, H.; Tomio, A.; Hirose, K.; Yutaka, K.; Yamamoto, Y. Validity, reproducibility, and clinical significance of noninvasive brachial-ankle pulse wave velocity measurement. Hypertens. Res. 2002, 25 (3), 359−364. (8) Hansen, T. W.; Staessen, J. A.; Torp-Pedersen, C.; Rasmussen, S.; Thijs, L.; Ibsen, H.; Jeppesen, J. Prognostic value of aortic pulse wave velocity as index of arterial stiffness in the general population. Circulation 2006, 113 (5), 664−670. (9) Nichols, W. W. Clinical measurement of arterial stiffness obtained from noninvasive pressure waveforms. Am. J. Hypertens. 2005, 18 (S1), 3S−10S. (10) Xu, L.; Meng, M. Q.-H.; Shi, C.; Wang, K.; Li, N. Quantitative analyses of pulse images in Traditional Chinese Medicine. Medical Acupuncture 2008, 20 (3), 175−189. (11) Akiyama, M.; Morofuji, Y.; Kamohara, T.; Nishikubo, K.; Tsubai, M.; Fukuda, O.; Ueno, N. Flexible piezoelectric pressure 973
DOI: 10.1021/acssensors.7b00230 ACS Sens. 2017, 2, 967−974
Article
ACS Sensors property relationships for electronics and optoelectronics. Appl. Phys. Rev. 2017, 4 (2), 021306. (30) Yang, T.; Xie, D.; Li, Z.; Zhu, H. Recent advances in wearable tactile sensors: Materials, sensing mechanisms, and device performance. Mater. Sci. Eng., R 2017, 115, 1−37. (31) Liu, S. B.; Wu, X.; Zhang, D. D.; Guo, C. W.; Wang, P.; Hu, W. D.; Li, X. M.; Zhou, X. F.; Xu, H. J.; Luo, C.; Zhang, J.; Chu, J. H. Ultrafast dynamic pressure sensors based on graphene hybrid structure. ACS Appl. Mater. Interfaces 2017, DOI: 10.1021/acsami.7b07311. (32) Liu, Q.; Chen, J.; Li, Y.; Shi, G. High-Performance Strain Sensors with Fish-Scale-Like Graphene-Sensing Layers for Full-Range Detection of Human Motions. ACS Nano 2016, 10 (8), 7901−7906. (33) Boland, C. S.; Khan, U.; Ryan, G.; Barwich, S.; Charifou, R.; Harvey, A.; Backes, C.; Li, Z.; Ferreira, M. S.; Möbius, M. E.; et al. Sensitive electromechanical sensors using viscoelastic graphenepolymer nanocomposites. Science 2016, 354 (6317), 1257−1260. (34) Li, X.; Zhang, R.; Yu, W.; Wang, K.; Wei, J.; Wu, D.; Cao, A.; Li, Z.; Cheng, Y.; Zheng, Q.; et al. Stretchable and highly sensitive graphene-on-polymer strain sensors. Sci. Rep. 2012, 2, 870. (35) Li, X.; Sun, P.; Fan, L.; Zhu, M.; Wang, K.; Zhong, M.; Wei, J.; Wu, D.; Cheng, Y.; Zhu, H. Multifunctional graphene woven fabrics. Sci. Rep. 2012, 2, 395. (36) Yang, T.; Wang, W.; Zhang, H.; Li, X.; Shi, J.; He, Y.; Zheng, Q.s.; Li, Z.; Zhu, H. Tactile sensing system based on arrays of graphene woven microfabrics: electromechanical behavior and electronic skin application. ACS Nano 2015, 9 (11), 10867−10875. (37) Mei, H.; Huang, R.; Chung, J. Y.; Stafford, C. M.; Yu, H.-H. Buckling modes of elastic thin films on elastic substrates. Appl. Phys. Lett. 2007, 90 (15), 151902. (38) Yang, T. T.; Zhong, Y. J.; Tao, D. S.; Li, X. M.; Zang, X. B.; Lin, S. Y.; Jiang, X.; Li, Z. H.; Zhu, H. W. Integration of graphene sensor with electrochromic device on modulus-gradient polymer for instantaneous strain visualization. 2D Mater. 2017, in press. (39) Luo, N.; Dai, W.; Li, C.; Zhou, Z.; Lu, L.; Poon, C. C.; Chen, S. C.; Zhang, Y.; Zhao, N. Flexible piezoresistive sensor patch enabling ultralow power cuffless blood pressure measurement. Adv. Funct. Mater. 2016, 26 (8), 1178−1187. (40) Luo, Z.; Zhang, S.; Yang, W.; Yang, Z. A research on characteristic information of pulse wave. J. Beijing Polytech. Univ 1996, 3, 71−79. (41) Li, B. N.; Dong, M. C.; Mang, V.; Un, M. P. An intelligent mobile cardiovascular monitoring device based on pulse wave. International Conference on Intelligent Sensing and Information Processing. IEEE 2005, 463−468. (42) O’rourke, M. F.; Pauca, A. L. Augmentation of the aortic and central arterial pressure waveform. Blood press. monit. 2004, 9 (4), 179−185. (43) Wu, Q.; Ma, Z.; Sun, Y. Noninvasive power spectrum analysis of radial pressure waveform for assessment of vascular system. J. Mech. Med. Biol. 2012, 12 (01), 1250021. (44) Lee, X.; Yang, T.; Li, X.; Zhang, R.; Zhu, M.; Zhang, H.; Xie, D.; Wei, J.; Zhong, M.; Wang, K.; et al. Flexible graphene woven fabrics for touch sensing. Appl. Phys. Lett. 2013, 102 (16), 163117.
974
DOI: 10.1021/acssensors.7b00230 ACS Sens. 2017, 2, 967−974
