Highly Sensitive Strain Sensor Based on Microfiber Coupler for Wearable Photonics Healthcare

Flexible strain sensors are essential components of wearable smart devices that perceive and respond to strain stimulations. However, the sensitivity and response time of most sensors require further improvement to detect subtle strains related to human bodies. Herein, an ultrasensitive flexible optical sensor with fast response time has been built based on a microfiber coupler encapsulated in polydimethylsiloxane. Benefiting from dramatic change of coupling ratio of the microfiber coupler under weak strain, this flexible strain sensor exhibits ultrahigh strain sensitivity (gauge factor, GF = 900), low detection limit (0.001%), ultrafast response time (<0.167 ms), wide sensing range (0.45%), and superior durability and stability (>10 000 cycles). Real‐time capturing and recognizing of respiration, broadband sound signals, and pulse waves at different sites of human body have been demonstrated based on this highly sensitive microfiber coupler sensor. Moreover, simultaneous detection of the wrist pulse and human voice has been achieved based on the frequency division multiplexing technology. This flexible photonics strain sensor could serve as the prototype of ultrasensitive flexible optical sensors with fast response time for the development of high performance and wearable healthcare devices.

Flexible strain sensors are essential components of wearable smart devices that perceive and respond to strain stimulations. However, the sensitivity and response time of most sensors require further improvement to detect subtle strains related to human bodies. Herein, an ultrasensitive flexible optical sensor with fast response time has been built based on a microfiber coupler encapsulated in polydimethylsiloxane. Benefiting from dramatic change of coupling ratio of the microfiber coupler under weak strain, this flexible strain sensor exhibits ultrahigh strain sensitivity (gauge factor, GF ¼ 900), low detection limit (0.001%), ultrafast response time (<0.167 ms), wide sensing range (0.45%), and superior durability and stability (>10 000 cycles). Real-time capturing and recognizing of respiration, broadband sound signals, and pulse waves at different sites of human body have been demonstrated based on this highly sensitive microfiber coupler sensor. Moreover, simultaneous detection of the wrist pulse and human voice has been achieved based on the frequency division multiplexing technology. This flexible photonics strain sensor could serve as the prototype of ultrasensitive flexible optical sensors with fast response time for the development of high performance and wearable healthcare devices. electric safety, and electromagnetic interference immunity. At present, flexible optical fiber strain sensors mainly realize the measurement based on an optical power detection scheme, including the mechanoluminescent effect, the optical loss effect, and the optical interference effect. Sensors based on the mechanoluminescent effect are to generate and convert mechanical stimuli-related strain into fluorescence signals based on intelligent mechanoluminescent materials. Liang [29] proposed a selfpowered stretchable strain sensor based on the integration of mechanoluminescent phosphors with an elastomer optical fiber and demonstrated a relatively high detection limit (10%) due to the high threshold pressure of ZnS:Cu mechanoluminescent materials. Sensors based on the optical loss are to measure the changes in transmitted light intensity as the fiber is deformed by strain, where the sensitivity is mainly determined by the compromise between the attenuation coefficient and transmission characteristics of optical sensing fiber. Zhang et al. [30] demonstrated a polydimethylsiloxane (PDMS)-encapsulated microfiber pressure sensor based on the transition from guided modes into radiation modes, which realized a low detection limit of 7 mPa. Zhang et al. [31] proposed a U-shaped microfiber sensor based on a bending-dependent transmittance variation, which realized the highly sensitivity airflow measurement with a resolution of 0.012 m s À1 . Zhu et al. [32] designed a self-assembled wavy optical microfiber strain sensor by detecting the microbending loss of the microfiber and achieved a detection limit of 0.5%. Pan et al. [33] proposed a pre-bent microfiber sensor for controlling the transition of the radiation modes into guided modes in the bending microfiber and achieved a detection limit of 0.25%. Guo et al. [34] proposed a flexible polymer fiber strain sensor based on the localized surface plasmon resonance absorption and scattering effect of gold nanoparticles, which showed a detection limit of 0.09%. Sensors based on the optical interference is to detect the optical power variation based on the strain-induced optical interference effect. Peng et al. [35] proposed a hybrid plasmonic microfiber knot resonator that weak strain would shift resonant wavelength resulting in output light power change and achieved a detection limit of 0.01%. Yu et al. [36] developed a flexible sensor based on evanescently coupled microfibers that are encapsulated in PDMS film achieved a strain detection resolution of 0.0012%. It is worth emphasizing that the interference effect in traditional silica optical fiber has been shown to measure strain with a sensitivity limit at sub-pε level, [37] which offers great enormous potentialities for improving the sensitivity and detection limit of flexible optical fiber strain sensors.
In this article, we propose a flexible optical fiber strain sensor with an ultralow detection limit of 0.001% strain based on microfiber coupler encapsulated in PDMS film. The strong coupling effect of this microfiber coupler sensor is highly sensitive to strain-induced coupling length change, [38] so the weak strain demodulation is achieved by using the coupling ratio (CR) between the two output fibers. The proposed CR demodulation can effectively resist the interference of hybrid variables, such as optical loss and input light intensity fluctuation. Based on these advantages, this flexible strain sensor exhibits ultrahigh strain sensitivity (gauge factor, GF ¼ 900), low detection limit (0.001%), ultrafast response time (<0.167 ms), wide sensing range (0.45%), and superior durability and stability (>10 000 cycles). These superior sensing properties make this flexible microfiber coupler sensor feasible for wearable photonics healthcare applications, such as monitoring ultraweak strain signals arising from a human's arterial pulse, detecting dynamic impacts associated with ant crawling, and restoring sound signals caused by weak vibrations of human voice. In addition, the simultaneous monitoring of arterial pulse signals and human voice has been achieved based on the frequency division multiplexing demodulation technique. This proposed flexible microfiber coupler sensor with high sensitivity and fast response shows great potential for human-machine interfaces, artificial intelligence, and soft robotics.

Design and Fabrication
The microfiber coupler is at the heart of our flexible sensor, and slight deformations of the coupler induced by external stimuli would cause visible variations in CR. The microfiber coupler was fabricated by tapering and fusing two standard telecommunication single-mode optical fibers (SMF-28, Corning) at the same time using a method known as the flame brushing technique, [39] as shown in Figure 1a. A hydrogen gas flow controller (CS200, Sevenstar) and high-precision translation stages (ESP301, Newport) were used to fabricate the required diameter of the microfiber coupler (refer to Supporting Information for more details on flexible microfiber coupler sensor preparation). For real-time monitoring of the microfiber coupler fabrication, a 1550 nm-wavelength light was coupled into one fiber, and two photodetectors (Thorlabs PDA50B-EC) were used to measure the transmitted power from the fibers. The fabricated microfiber coupler had a low insertion loss of about 0.05 dB. The second step was to encapsulate the microfiber coupler in a flexible sandwich structure to improve its stability. Because of its high flexibility, transparency, and biocompatibility, PDMS film was chosen as the encapsulation layer. The PDMS precursor (mixing ratio of the base polymer and curing agent) was poured into a homemade model groove (2 cm Â 5 cm size with 100 μm thickness) as the bottom layer through thermal treatment at 80°C for 30 min to obtain a flat and smooth film. The prepared bottom PDMS film was transferred to the other deeper groove mold (2 cm Â 5 cm size with 250 μm thickness) for the assembling process. After that, the fabricated microfiber coupler was transferred onto the bottom PDMS film in the deeper groove mold, where it could firmly adhere due to van der Waals forces. Finally, after 30 min of curing at 80°C, some PDMS precursor was uniformly dispersed in the deeper groove mold to cover the microfiber coupler with a thickness of 150 μm, and the layered flexible microfiber coupler sensor was fully assembled and realized. After encapsulation, alcohol could be used to help reduce the adsorption capacity and enable nondestructive separation of the PDMS flexible sensor from the groove mold. A highly transparent sensor film with a thickness of about 250 μm was demonstrated to manipulate in twisted ( Figure 1c) and bent states (Figure 1d,e), demonstrating exceptional structural flexibility. Figure 1b,e showed that even after massive deformations, this encapsulated microfiber coupler maintained a good light-guiding ability. This encapsulated microfiber coupler was stuck on the wrist by van der Waals force with good contact, indicating that the sensor had good wearability, as shown in Figure 1f. This prepared flexible sensor with skin-like mechanical properties and excellent optical transparency made it a potent tool for wearable applications.

Results and Discussion
As shown in Figure 2a, a distributed feedback (DFB) laser diode (FITEL, Furukawa Electric Co.) operating at 1550 nm with a 1.1 MHz linewidth was used as a light source, and two photodetectors (Thorlabs PDA50B-EC) were used as the output power meters. Here, the CR (the CR referred to the ratio of P 3 to the total output power (P 3 þ P 4 ), CR ¼ P 3 P 3 þP 4 ) was set as the sensing parameter, which could be theoretically analyzed according to coupled mode theory. [40] CR ¼ sin 2 ðκlÞ where κ and l were the coupling coefficient and the coupling length, respectively. κ was determined by the microfiber radius r, the refractive index of silica glass n 1 , and the PDMS n 2 . When external strain ε was applied to the microfiber coupler, both κ and l changed. The coupling coefficient change Δκ was determined by the microfiber radius r, the refractive index of silica glass n 1 þ Δn 1 , and the PDMS n 2 þ Δn 2 under external strain ε. The change in refractive index was given by where v, P 11 , and P 12 were the Poisson's ratio and elastic coefficients, respectively. [41,42] The elongation Δl was εl.
When Δκ and l were small enough, the change in CR ΔCR was given approximately by Based on Equation (1-3), external strain ε could be calculated by analyzing the ΔCR of the microfiber coupler. For example, Figure 2b showed the CR of a microfiber coupler (r ¼ 1 μm) under 0%-0.005% strain. When the coupling length was 18.125 mm (CR ¼ 0.5 with ε ¼ 0), the CR decreased linearly with increasing strain.
The strain sensitivity was evaluated by the gauge factor (GF). By substituting Equation (3), the GF was Based on Equation (4), the strain sensitivity GF had been expressed as the product of two terms: the envelope function κ þ Δκ ε À Á l. and the sinusoidal function |sin(2κl)|. To obtain a higher sensitivity, both two terms should be considered. The first term was monotone increasing as l and κ increased. Considering that the κ decreased as the microfiber radius r increased, [43] the sensor would be more sensitive with a thinner microfiber and a longer coupling length. The second term was periodic and reached a maximum only when 2κl ¼ 2m AE 1 2 À Á π, where m was an integer. This term indicated that the microfiber coupler sensor must have an optimal operating point. By substituting (1), CR ¼ 0.5 had been confirmed for the initial optimal operating point, which could be achieved by prestretching of the microfiber coupler. Figure 2c showed the GF calculations with different radii and coupling lengths under the 0.001% strain. When l > 20 mm and r < 0.8 μm, the theoretical value of GF could be larger than 2000, which was greater than the theoretical value of most reported flexible strain sensors.
The strain sensing performance of our proposed flexible microfiber coupler sensor was characterized by a tensile strain tester. Under a strain range of 0%-0.07%, the sensors with the microfiber coupler radii of 1.0 μm, 1.5 μm, and 2.0 μm were stretched. As shown in Figure 2d, the 1 μm-radius microfiber coupler sensor had the highest sensitivity and relatively good linearity. Due to the low refractive index contrast between the silica microfiber coupler and PDMS encapsulation material, the www.advancedsciencenews.com www.advintellsyst.com microfiber coupler sensor with a radius of less than 1 μm would suffer from high optical loss and low detectable signal. To investigate the weak strain sensing performance, the 1 μm-radius microfiber coupler sensor was further stretched and investigated with a step of 0.001% strain. Figure 2e showed an ultrahigh sensitivity of GF ¼ 900 according to the calculated slope, which was www.advancedsciencenews.com www.advintellsyst.com consistent with our theoretical analysis as the red star marker position illustrated in Figure 2c. Furthermore, limited by the tensile strain tester's resolution, the minimum strain and strain step applied to this sensor was only 0.001%, and the excellent linearity and obvious separation of the experimental curve points shown in Figure 2e indicated that the actual detection limit of this microfiber coupler sensor would be much lower than 0.001%. The blue curve in Figure 2f illustrates how the strain sensing range would be constrained inside a quarter cycle of a strain-CR cosine curve, or just 0.034%, if the microfibre coupler sensor operated at 0.001% detection limit scenario. If the sensing range was expanded, then the detection limit would be sacrificed. As shown in Figure 2d, compared to the r ¼ 1.5 μm and r ¼ 2 μm situations, the sensitivity GF would be decreased from 900 to 304 and 190, respectively. To solve the paradox of low detection limit and broad sensing range, the optical loss of the microfiber coupler under applied strain was used as the second measurand with CR simultaneously. Here, the optical loss referred to the ratio of the total output power (P 3 þ P 4 ) to the input optical power P 1 , αðdBÞ ¼ À10lg P 3 þP 4 P 1 . As indicated by the red curve in Figure 2f, the optical loss had a monotonically increasing trend up until the applied strain reached 0.45%. We discovered that varied optical loss values could precisely identify the period of the strain-CR cosine periodic functional curve by combining the cosine periodic functional strain-CR response and monotonically increasing strain-optical loss response simultaneously. In addition, the method of one-to-one corresponding optical loss and CR could clearly locate the strain state of the sensor. Due to the hysteresis effect of PDMS material, the optical loss values did not display a strictly linear data state. Actually, it was precise because such inevitable slight fluctuations in the strain-optical loss curve, these optical loss effect-based flexible strain sensors [30][31][32][33][34] could not achieve low detection limits and high sensitivity. However, these slight fluctuations in the strainoptical loss curve did not affect the practical expansion of the strain sensing range in our work. It is worth noting that the demodulation of the CR and loss variation are also capable of simultaneous measurement of strains and other sensing parameters, such as temperature, which is beyond the scope of this article. The temperature was kept constant during the strain test, and the cross sensitivity of the strain and temperature were ignored.
To further investigate the stability and repeatability of this optical microfiber coupler strain sensor, a periodic stretchingreleasing motion was applied to the sensor with a fixed 0.12% strain for more than 10 000 cycles. As shown in Figure 2g, the CR variation amplitude was stable throughout the test. The detailed plots of the areas outlined in red, green, and blue (corresponding to the earlier, middle, and later stages of the cycle testing, respectively) were also shown in the upper magnified inset of Figure 2g, where the strain-induced CR variation curves remained constant throughout the testing.
The response time of this flexible microfiber coupler sensor had been thoroughly investigated. The tensile strain tester's maximum stretching frequency was capped at Hz-scale, which was substantially lower than the sensor's frequency response bandwidth. Here, we used a noncontact acoustic pressure testing approach for the high-speed and ultralow strain application. A loudspeaker driven by a signal generator was used as a dynamic strain-applying source and placed 10 cm away from the sensor film. The acoustic pressure level of the 3 kHz sinusoidal sound wave produced by the loudspeaker was 83 dBC, as determined by a calibrated electronic microphone placed close to the sensor film. We obtained the CR change time-varying waveform, as shown in Figure 2h, in which a nearly undistorted sinusoidal shape could be clearly distinguished. As shown in Figure 2i, the frequency spectrum of the CR change waveform was analyzed, and the global SNR of the applied 3 kHz frequency component was estimated to be around 24 dB. According to the Nyquist-Shannon sampling theorem, [44] the response frequency of our flexible microfiber coupler sensor was reasonable to assume at least twice the applied 3 kHz. So the response time could be calculated as Compared with previously reported high-sensitivity flexible strain sensors, the detection limit of our proposed microfiber coupler sensor had reached the top level, and the sensitivity had exceeded the highest level among flexible optical strain sensors, as shown in Table 1.
Owing to its superior mechanical properties and high sensitivity, this proposed microfiber coupler sensor has enormous potential applications in wearable devices for monitoring ultraweak physiological signals, slight human motions, and subtle environmental perturbations. The microfiber coupler strain sensor was developed using PDMS as an encapsulation material to be a biocompatible, flexible, and durable device that could be easily and firmly attached to the human skin while remaining  Figure 3a, this flexible microfiber coupler sensor could be directly attached to the skin of the human face, neck, wrist, finger, and ankle, for real-time detection of breath, pulse, gesture, and speech. Breathing is one of the most important functions performed by humans, and it has significant impact on morphology, craniofacial and cervical functions, and is also linked to some respiratory diseases. This flexible microfiber coupler strain sensor was mounted on facial www.advancedsciencenews.com www.advintellsyst.com masks to monitor strain caused by respiratory motions, where inhalation and exhalation would bring about air pressure changes inside the mask and deform it. The distinguishable response patterns in Figure 3b were used to detect and discriminate three different respiration modes: 0.2 ΔCR and 38 times min À1 for deep breath, 0.07 ΔCR and 23 times min À1 for normal breath, and 0.02 ΔCR and 17 times min À1 for shallow breath. The respiration frequency, breathing depth, and various breathing styles could be sensed and detected by detecting the mask deformation, which would be helpful to discover and diagnose some respiratory diseases. Arterial pulse is a significant physiological signal for clinical diagnosis of cardiovascular diseases. The arterial pulse is assessed for the contour of the pulse wave as well as its volume, rate, and rhythm; however, the intensity of the arterial pulse signal is frequently too weak to be palpated or detected, particularly at the fingertip and ankle sites. This flexible microfiber coupler strain sensor was attached to different body sites, such as the neck, wrist, fingertip, and ankle. By virtue of high sensitivity and low detection limit, this flexible strain sensor can precisely detect and record the pulse waveforms at each part of the body, as shown in Figure 3c. The pulse waveform details could be captured and recovered without distortion; for example, the pulse signals at the neck, wrist, and finger had three peaks, while the ankle only had two peaks, which was exactly in agreement with the pulse signal characteristics at the body's various sites. [13] The weak motions of the human body could be effectively monitored by this flexible microfiber coupler sensor, which would have broad application prospects in human-machine interaction and phonation rehabilitation training. First, this flexible microfiber coupler sensor was attached to the wrist site for the gesture recognition demonstration, where finger bending would cause the wrist to produce weak movements. As shown in Figure 3d, the bending and straightening of different fingers could be clearly recognized and distinguished. Then, this flexible microfiber coupler sensor was attached to the neck to monitor the tiny epidermis and muscle movements during speech for phonation recognition. As shown in Figure 3e, this sensor captured distinguishable and repeatable signal patterns when the volunteer spoke some alphabet letters, S-C-U-T. The three enlarged detail views of letter "T" waveforms were found to be nearly identical, indicating that this flexible sensor would be more powerful for in situ real-time monitoring based on additional pattern recognition techniques.
To demonstrate that this flexible sensor had the characteristics of fast response, high sensitivity, and continuous monitoring, an experiment was designed and conducted to monitor the ant-crawling-induced dynamic weak strains. The flexible microfiber coupler sensor was mounted on a stage with stretched form, and an ant weighing 1.1 mg could freely crawl from left to right across the sensor film surface (Figure 4a). The sensor had captured the ant-crawling-induced weak strain signals in real time during the crawling movements (Movie 1 in Supporting Information). The recorded parabolic shape curve, as shown in Figure 4b, indicated that this sensor was completely capable of detecting ultralight ant movement. However, the nonflat response curve revealed that the sensitivities were not distributed uniformly on the sensor film, thereby causing difficulty in quantitative dynamic measurement.
To further explore the dynamic weak strain sensing ability, this flexible microfiber coupler sensor was tested as a noncontact acoustic microphone. As shown in Figure 4c, the sensor was stretched and suspended above a loudspeaker driven by a signal generator. The loudspeaker generated single-frequency sinusoidal sound waves ranging from 20 Hz to 20 kHz, with increments of 10 Hz, 100 Hz, and 10 kHz. The acoustic pressure level was set to around 96 dBC at all of these acoustic-strain frequency levels, as measured by a calibrated electronic microphone. As illustrated in Figure 4d, the measured CR under logarithm function obviously response to the acoustic pressure-induced strain, where the 3 dB frequency response bandwidth was mainly located within 50 Hz-3 kHz and the ΔCR decayed rapidly at frequencies lower than 50 Hz and higher than 3 kHz. The main reason was that the frequency response range, particularly the high-frequency portion, was significantly degraded by the PDMS film's flexibility. As illustrated in Supporting Information (Figure S1), the recorded CR at the 300 Hz and 3 kHz frequency testing points could be clearly captured and recognized with a nearly undistorted sinusoidal shape. According to the amplitude of CR change, the strain induced by the acoustic wave pressure on the sensor was approximately 0.000011%, which was much lower than the previously measured detection limit of 0.001%. In contrast to the single-frequency sinusoidal acoustic strain used in the preceding experiments, the real human voice is a complex signal composed of acoustic waves of varying frequencies and intensities that changed rapidly over time. To verify the microphone function of this flexible sensor, an audio clip (music "You raise me up," 15 s duration) was applied to the loudspeaker and broadcasted to the sensor. Figure 4e depicts the demodulated CR change frequency spectrum under the music audio clip source, which was mainly concentrated in the 100 Hz-3 kHz range, with an inset showing the CR change time-varying waveform. This time-varying demodulated CR change waveform was converted into an audio file (Audio 1 in Supporting Information). The music could be clearly recognized with little noise when listening to this demodulated audio file, demonstrating the fidelity of the flexible microphone application.
Additionally, a novel frequency division multiplex-based mechanism for multiparameter sensing was made possible by this sensor's broadband response. As shown in Figure 4g, the flexible sensor was mounted to the skin of the wrist, so the arterial pulse signals from the attached wrist and the human voice signals broadcasted from the loudspeaker could apply simultaneously to this sensor. Because arterial pulse signals were mostly in the 0.3-2 Hz frequency range and human voice signals were in the 100 Hz-3 kHz frequency range, these two signals could be monitored and demodulated simultaneously by this flexible sensor (Figure 4f ). As shown in Figure 4g, the record signals were enveloping curves of low-frequency pulse modulated with high-frequency sound. The pulse (75 times per minute) and the audio signals were individually filtered and presented using frequency domain filter techniques. The restored sound was similar to the original audio, as evidenced by Audio 2 in Supporting Information, thus validating the high fidelity of the sensor.

Conclusion
In summary, we have developed a flexible strain sensor with high sensitivity and fast response using a PDMS film-encapsulated optical microfiber coupler structure. The CR between the two output fibers of the microfiber coupler has been used to realize weak strain demodulation. This flexible strain sensor exhibits ultrahigh strain sensitivity (gauge factor, GF ¼ 900), low detection limit (0.001%), ultrafast response time (<0.167 ms), wide sensing range (0.45%), and superior durability and stability (>10 000 cycles). These superior sensing properties make this flexible microfiber coupler sensor feasible for wearable photonics healthcare applications, such as the accurate detection of the arterial pulse, ant crawling, and human voice. This proposed flexible microfiber coupler sensor paves the way for future wearable optical devices for highly sensitive sensor applications including personalized monitoring of the cardiovascular systems, human-machine interfaces, soft robotics, and artificial intelligence.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author. c) The setup of the microphone application. d) Frequency response characteristics of this flexible microfiber sensor. e) The frequency spectrum of the recorded CR change induced by the music broadcast. The inset is the related CR time-varying waveform. f ) Frequency division multiplexing mechanism based on the proposed strain sensor. g) Simultaneously monitoring and demodulating arterial pulse and human voice signals using this proposed wrist-mounted sensor.