Highly Stretchable PPy/PDMS Strain Sensors Fabricated with Multi-Step Oxygen Plasma Treatment

We present highly stretchable polypyrrole (PPy)/polydimethylsiloxane strain sensors of highly improved sensitivity and durability fabricated by a chemical oxidative polymerization with oxygen plasma treatment (O2 PT). In this study, O2 PT was performed for 30, 60, and 90 s at each growth stage of the PPy film in three steps to investigate the effects on the sensor performance as well as the microstructural properties of the PPy films. Bonding characteristics with underlying layers and resistance to microcrack generation of the multi-layer PPy films under our given strained state were significantly enhanced by the O2 PT. The best sensor performance in terms of sensitivity and stability were achieved by PT for 30 s with a maximum gauge factor of ~438 at a uniaxial strain of 50%, excellent durability over 500 stretching/release cycles, and a fast response time of ~50 ms.


Introduction
Stretchable strain sensors have gained great recognition in terms of their potential in applications such as continuous health monitoring [1], human body motions [2][3][4], humanto-machine interfaces [5], and intelligent robots [6] with significant research advances in wearable technology. In general, they can be classified into various categories of capacitive [7], resistive [8,9], triboelectric [10,11], and piezoelectric [12,13] sensors depending on their operation mechanism. Among all of these sensing devices, the resistive strain sensors have been the most preferred and intensively studied because of their simple structure, user-friendly and secure signal acquisition, and high sensing stability [14][15][16]. Nonetheless, it is still a great challenge to successfully fabricate resistive strain sensors capable of enduring a wide strain range (greater than few tens in percentage), high gauge factor, long-term durability, and good interoperability with body skin.
To meet all these demanding conditions, many different stretchable conducting polymer materials of polypyrrole (PPy) [17][18][19], polyaniline (PANI) [20][21][22], and poly (3,4ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) [23,24] have been intensively studied for their easy integration to flexible and human-skin friendly substrates such as polydimethylsiloxane (PDMS), thermoplastic polyurethane (TPU), and Ecoflex TM rubbers [25][26][27]. More specifically, these conducting polymers, which have recently started to be widely examined as strain sensor films, are subjected to respond to mechanical deformations which, in turn, alters their electrical characteristics, such as change in resistance, because of their reproducibility and stretchability. In particular, strain sensors based on these stretchable conducting polymers have great importance in the everyday health requirements of people which include the tracking of real-time fitness, routine health monitoring, and, more importantly, the diagnostics of early diseases [28]. Among all the properties of stretchable conducting polymers films for strain sensors, stretchability, conductivity, and sensitivity are three fundamental factors which have decisive roles in the

Materials and Methods
The pyrrole (C 4 H 5 N, 99.5% in purity) and iron chloride (FeCl 3 , 97% in purity) were from Sigma-Aldrich, South Korea and used without any further purification. N-(3-trimethox ysilylpropyl) pyrrole (C 10 H 19 NO 3 Si, 95% in purity), abbreviated as Py-Silane hereafter, was from ABCR GmbH, Karlsruhe, Germany. The SYLGARD TM 184 kit for silicone elastomer base and curing agent were purchased from Dow Corning Corporation (Auburn, MI, USA). Colloidal silver paste (PELCO ® , particle size 0.4-1.0 µm) purchased from Ted Pella, Inc. (Redding, CA, USA) was used for the electrode formation of sensor devices at low temperature. First, thin PDMS elastomer substrates were prepared by mixing the silicone elastomer base and curing agent in a volume ratio of 10:1. The liquid mixture was then placed on a watch glass and, after degassing, transferred to a silicon wafer by spin-coating at 200 rpm for 30 s. The spin-coated PDMS was finally cured at 100 • C for 2 h to achieve a final thickness of 0.45 mm and tailored to a rectangular shape (8 × 10 mm) using a surgical blade for further use.
Schematics of Figure 1a,b describe a chemical oxidation polymerization procedure with three-step O 2 PT to fabricate our multi-layer PPy/PDMS strain sensor structures. In the first step, the PDMS substrate was plasma-treated with O 2 in a reactive ion etching system for 3 min at a RF power of 50 W, O 2 flow rate of 100 sccm, and a process pressure of 50 mtorr to improve the PPy adhesion to the substrate by enhancing the hydrophilicity. The contact angle (CA) was~103.7 • before PT, and it became significantly smaller (~29.3 • ) after PT on the PDMS surface. CA measurements were done by a tensiometer (DAS100, Kruss, Germany) with an optical goniometer using the water sessile drop method. As illustrated in Figure 1, this O 2 PT resulted in the formation of highly populated silanol groups (Si-OH) on the PDMS substrate and lead to the substrate surface being more reactive to the silane groups such as Py-Silane, used for a seed layer in this study. The PDMS substrate was then placed in a vial glass with a drop of~5 µL Py-Silane near to it, and the setup was kept in a vacuum oven and heated at 70 • C for 3 h. During this vacuum vapor treatment of PDMS surface by Py-Silane, the silane components of the Py-Silane monolayer are to be coupled with the hydroxyls (-OH) on PDMS as shown in Figure 1b. Our chemical oxidation polymerization process was carried out using FeCl 3 as an oxidant, and it lead to the pyrrole monomers being subjected to surface-initiated polymerization. PPy polymerization was carried out by dipping the O 2 plasma-treated PDMS substrates into the growth solution mixture (for 30 min) which was prepared by mixing the aqueous solutions of FeCl 3 (0.5 M/12 mL) and pyrrole (1 M/12 mL). During the polymerization process, an interaction between the Py part (see Figure 1b) of Py-Silane with Py-monomer takes place, and it creates the strong covalent bonds of PPy on the PDMS surface as shown in Figure 1b. After the first-step growth of PPy films, the samples were taken out from the growth solution and rinsed in deionized water with ultrasonic agitation (for 3-4 min) to remove any loosely coupled polymers on the surface. Microstructural properties of PPy films on PDMS prepared under different O2 PT conditions were characterized by a field emission scanning electron microscope (FESEM, S-4800, Hitachi, Japan) at 15 kV and optical microscopy (BH3-WHP6, Japan). Electrical resistances of the fabricated sensor devices were measured at room temperature through the electrodes formed on the PPy surface using a Keithley 2450 source meter unit. Normalized resistance change (ΔR/Ro) upon static and cyclic uniaxial stretch/release for the sensor devices was measured by a universal tensile test machine (Digital Tech, Japan), The PPy layer grown on PDMS by the first step is shown in a photograph of Figure 1c. Before the second-step growth, the films grown in the first step were plasma-treated with O 2 as shown in the second row of Figure 1a. When the PPy films are grown by this multi-step method with O 2 PT, oxygen radicals present in the vacuum play an important role in changing the PPy micro-fibers to a stronger and more densely cross-linked structure bound more tightly to the underlying layers. After the second-step growth in the growth solution for 30 min as shown in Figure 1d, third-step growth was performed in the same way as illustrated in the third row of Figure 1a. The third layer of PPy film synthesized in this step is shown in a photograph of Figure 1e. A lower RF power of 15 W but the same O 2 flow rate (100 sccm) and process pressure (50 mtorr) were used for the O 2 PT in the second and third steps to minimize the plasma damage on the PPy films. Multi-layer PPy/PDMS strain sensors with three different PT time intervals of 30, 60, and 90 s were fabricated, where the control sample (CS) with no treatment (PT for 0 s) was also prepared to investigate the effect of oxygen plasma exposure to the sensor performance. The PPy films synthesized in this procedure showed the total thickness in a range of 500 ± 30 nm as measured by cross-sectional SEM. Electrodes for each sensor element were finally formed using silver paste at room temperature and wired for further measurement as shown in Figure 1f.
Microstructural properties of PPy films on PDMS prepared under different O 2 PT conditions were characterized by a field emission scanning electron microscope (FESEM, S-4800, Hitachi, Japan) at 15 kV and optical microscopy (BH3-WHP6, Japan). Electrical resistances of the fabricated sensor devices were measured at room temperature through the electrodes formed on the PPy surface using a Keithley 2450 source meter unit. Normalized resistance change (∆R/R o ) upon static and cyclic uniaxial stretch/release for the sensor devices was measured by a universal tensile test machine (Digital Tech, Japan), where R o is the unstrained resistance, and ∆R is the resistance change (∆R = R − R o ) upon uniaxial stretching. A uniaxial stretch/release test was performed at a strain rate of 100 mm/min, a tensile force of 5 N, and a clamp distance of 10 mm. One clear difference observed in SEM micrographs is that the PPy layers grown with O 2 PT exhibit a smoother surface with higher surface coverages as depicted in Figure 2e-p, while the PPy films prepared with no O 2 PT (control sample, CS) shows rougher surface (Figure 2a-d). This observation suggests that the O 2 PT significantly affects the bonding characteristics of additional PPys to the lower layers such as the PDMS or underlying PPys during the growth. It, in fact, provides rich hydroxyl groups to the surface of underlying polymer layers as well as more hydrophilic surface, thereby allowing the nucleation of PPys to covalently bond more actively with the underlying layer and ensuring the deposition of denser fiber networks [39]. For example, the earlier studies of O 2 PT or oxygen plasma immersion ion implantation methods showed that the static water CAs on the polymeric surfaces of PDMS, poly(ε-caprolactone), and PPy were decreased from 114, 136, and 98 • tõ 25, 0, and 3 • by their optimum conditions [40][41][42]. In our case, the CA was decreased from 123 to~37 • after PT on the PPy surface. Thin PPy films with rougher surface also showed more "peel-off" problems which can lead to increase in electrical resistance and lower responsivity upon applied external strain. On the other hand, uniform surface coverage and the formation of a dense micro-fiber network of PPy achievable by the interlayer O 2 PT is essential for reliable long-term strain sensing performance. polymeric surfaces of PDMS, poly(ε-caprolactone), and PPy were decreased from 114, 136, and 98° to ~25, 0, and 3° by their optimum conditions [40][41][42]. In our case, the CA was decreased from ~123 to ~37° after PT on the PPy surface. Thin PPy films with rougher surface also showed more "peel-off" problems which can lead to increase in electrical resistance and lower responsivity upon applied external strain. On the other hand, uniform surface coverage and the formation of a dense micro-fiber network of PPy achievable by the interlayer O2 PT is essential for reliable long-term strain sensing performance. Microcrack densities of the PPy films after being stressed by 500 cyclic uniaxial stretching (20%) and release were measured to quantitatively examine the effect of O 2 PT. A simple linear density to the elongational direction was utilized in this study because all the microcrack lines were aligned to the perpendicular direction with elongation as shown in Figure 3. Optical micrographs were taken at four different areas (50 × 50 µm 2 ) under a sustaining strain of 5% so that all the microcracks could be observed more clearly. Straight lines were drawn to the parallel direction to the elongation for each selected area on optical images, and the average numbers of microcracks intersecting the crossing lines were extracted. Figure 3b illustrates the extracted microcrack densities (number of microcracks per µm) of the multi-layer PPy/PDMS samples prepared under various O 2 PT conditions. The lowest density was shown in the case of O 2 PT for 30 s, while the highest was obtained from CS, and this result clearly shows that our O 2 PT (especially, the maximum effect at PT sustaining strain of 5 % so that all the microcracks could be observed more clearly. Straight lines were drawn to the parallel direction to the elongation for each selected area on optical images, and the average numbers of microcracks intersecting the crossing lines were extracted. Figure 3b illustrates the extracted microcrack densities (number of microcracks per μm) of the multi-layer PPy/PDMS samples prepared under various O2 PT conditions. The lowest density was shown in the case of O2 PT for 30 s, while the highest was obtained from CS, and this result clearly shows that our O2 PT (especially, the maximum effect at PT for 30 s) allows the PPys to grow to a denser fiber network structure, thereby resulting in enhanced resistance to microcrack generation under alternating stress.   30, 60, and 90 s. A total of 500 cyclic loads of stretching (20%)/release were applied in advance to the PPy films before optical microscope observation. Micrographs were taken on four different areas (50 × 50 µm 2 ) for each PT condition. (b) Bar-graphs illustrating the microcrack densities measured from the PPy films plasma-treated for different time intervals of 0 (CS), 30, 60, and 90 s. All the images were taken under the same uniaxial stretching condition (5%) to make the surface microcracks more visible.

Electrical Characterization
Shown in Figure 4a-d are the typical current-voltage (I-V) characteristics of multi-layer PPy/PDMS strain sensors fabricated using different O 2 PT conditions (CS, PT for 30, 60, and 90 s) when stressed under various uniaxial strains ranging from 0 to 50% elongation. Regardless of PT condition, all the sensors showed good ohmic properties in a voltage range from −5 to +5 V, but the sensor plasma-treated for 30 s exhibited the lowest electrical resistance (reciprocal of the slope in I-V curve). The electrical resistance progressively increased with the increase of strain in all cases, but the resistances in the case of PT for 30 s were the lowest over the entire range of strain as shown in Figure 4e. As observed in our SEM, the resistance force against microcrack generation of PPy film was maximum at a PT time interval of 30 s but decreased again when the PT was performed for longer time intervals. This phenomenon is not fully understood yet, but it is believed that the increased nucleation rate of PPys caused by higher population of hydroxyls (as well as hydrophilicity) on the surfaces of underlying polymers starts to be saturated to a level of O 2 plasma exposure time as reported in the previous studies [43]. On the other hand, for PT longer than 30 s, delocalization of the π-conjugated electrons from the backbone chains of PPys can be retarded by the introduction of negatively charged oxygen ions in the plasma state and highly populated hydroxyl groups on the surface; therefore, beyond this point, the intrinsic resistance force of PPy against microcrack generation can start to decrease. at a PT time interval of 30 s but decreased again when the PT was performed for longer time intervals. This phenomenon is not fully understood yet, but it is believed that the increased nucleation rate of PPys caused by higher population of hydroxyls (as well as hydrophilicity) on the surfaces of underlying polymers starts to be saturated to a level of O2 plasma exposure time as reported in the previous studies [43]. On the other hand, for PT longer than 30 s, delocalization of the π-conjugated electrons from the backbone chains of PPys can be retarded by the introduction of negatively charged oxygen ions in the plasma state and highly populated hydroxyl groups on the surface; therefore, beyond this point, the intrinsic resistance force of PPy against microcrack generation can start to decrease.  To further investigate the sensing performance, the gauge factors (GFs) of multi-layer PPy/PDMS structures prepared under different PT conditions were calculated. The GF is defined as (∆R/Ro)/ where is the mechanical strain, and ∆R/Ro is the normalized resistance change, and it is an important parameter demonstrating the sensitivity of resistive strain sensors. Plotted in Figure 5 are the calculated GFs as functions of uniaxial strain for the sensors fabricated under two different conditions (CS and PT for 30 s). The maximum GFs of each sensor prepared under every different PT condition at a strain of 50% ( = 0.5) were also summarized in an inset table. As shown in Figure 5a, the CS exhibited the lowest GF of 97 at 50% strain, and the reason behind this low sensitivity is the absence of PT which leads to the loosely bounded PPy layers and the consequent peel-off or break-down under strained condition. On the other hand, the GF values of the sensors prepared with O 2 PTs showed significant improvement, and, as illustrated in inset table of Figure 5a, the highest GF of 438 was obtained from the case of PT for 30 s (at = 0.5). Various GFs of stretchable sensor materials on PDMS have been reported such as 2-10 from Ag-nanowire/PDMS and carbon black nanoparticle/PDMS, 137-261 from graphene oxide/PDMS,~103 from graphene woven fabrics/PDMS, and 3.9-233 from graphene/PDMS composite [44]. In the case of a free-standing silver/PPy composite sensor, a GF of~21 was reported at a strain of 20% [45]. In addition, the comparison of previously reported flexible strain sensors based on PDMS were summarized in Table 1. The greater GF of a sensor we have, the more easily we can detect small deformations. One notes that a GF of 438 obtained in this study (multi-layer PPy/PDMS structures fabricated with PT for 30 s) at a very high strain of 50% is a significant achievement.  [44]. In the case of a free-standing silver/PPy composite sensor, a GF of ~21 was reported at a strain of 20% [45]. In addition, the comparison of previously reported flexible strain sensors based on PDMS were summarized in Table 1. The greater GF of a sensor we have, the more easily we can detect small deformations. One notes that a GF of 438 obtained in this study (multi-layer PPy/PDMS structures fabricated with PT for 30 s) at a very high strain of 50% is a significant achievement.   Strain load of a quasi-step function varying between 0 and 1% was applied to the sensors to examine the transient response characteristics. Magnified views of the measured single transient responses from the CS and the sensor plasma-treated for 30 s are, respectively, shown in Figure 5b and additional inset. Response times (τ RS ) for the CS and plasma-treated (for 30 s) sensors to reach their quasi-steady states were 200 and 50 ms, respectively, while τ RS of the sensors that were plasma-treated for 60 and 90 s were 130 and 140 ms, respectively (see the inset table). As revealed from this measurement, the O 2 PT for 30 s also brought about significantly reduced response time that can carry out real-time detection and respond to complicated human activities.
To investigate the sensing stability of our strain sensors with the change of strain in magnitude, the responses of ∆R/R o upon the cyclic stretching/release were measured as depicted in Figure 6 by increasing the peak elongational strain from 10 to 50% with a 10% stepwise increase. For these measurements, a constant tensile strain rate of 100 mm/min was maintained. Except for the case of PT for 30 s, the normalized resistances at either peak point (stretching)  In these cyclic tests, the peak strain was varied from 10 to 50% with a stepwise increase of 10%.
To examine the sensing stability depending upon the change of strain rate, the sensors were examined under another stretching/release cycle test condition with three different strain rates of 10, 50, and 100 mm/min after each set of three cycles, as illustrated in Figure 7a-d, but the maximum stretching strain was fixed at 20% in all cases. It was confirmed that the sensor with O2 PT for 30 s also exhibited no significant variation in normalized resistance and the stable response at different strain rates as shown in Figure 7b. The CS sensor, however, showed clear and gradual increase in normalized resistance, especially at peak points (stretching), with the increase of strain rate as revealed in Figure  7a. In this test, the sensors prepared with PT for 60 or 90 s exhibited much more stable responses in normalized resistance (see Figure 7c-d) than those of CS. To examine the sensing stability depending upon the change of strain rate, the sensors were examined under another stretching/release cycle test condition with three different strain rates of 10, 50, and 100 mm/min after each set of three cycles, as illustrated in Figure 7a-d, but the maximum stretching strain was fixed at 20% in all cases. It was confirmed that the sensor with O 2 PT for 30 s also exhibited no significant variation in normalized resistance and the stable response at different strain rates as shown in Figure 7b. The CS sensor, however, showed clear and gradual increase in normalized resistance, especially at peak points (stretching), with the increase of strain rate as revealed in Figure 7a. In this test, the sensors prepared with PT for 60 or 90 s exhibited much more stable responses in normalized resistance (see Figure 7c To assess the durability of multi-layer PPy/PDMS strain sensors, we also performed a mini-marathon test of 500 cyclic stretching/release in real time with a constant peak strain of 20% and a fixed strain rate of 100 mm/min. As shown in Figure 8a, the normalized resistances of CS exhibited significant increase with time at both peak and valley points, which could be attributed to the fatigue fracture and displacement of PPy microfiber structure loosely bounded to the underlying layer due to the absence of interlayered O2 PT. On the other hand, the sensors fabricated with O2 PT showed more stable peak and valley values and durable performance under this cyclic test as shown in Figure 8b-d. Especially, the sensor with PT for 30 s demonstrated the best performance in durability over the entire range of the cyclic test (Figure 8b). To assess the durability of multi-layer PPy/PDMS strain sensors, we also performed a mini-marathon test of 500 cyclic stretching/release in real time with a constant peak strain of 20% and a fixed strain rate of 100 mm/min. As shown in Figure 8a, the normalized resistances of CS exhibited significant increase with time at both peak and valley points, which could be attributed to the fatigue fracture and displacement of PPy microfiber structure loosely bounded to the underlying layer due to the absence of interlayered O 2 PT. On the other hand, the sensors fabricated with O 2 PT showed more stable peak and valley values and durable performance under this cyclic test as shown in Figure 8b-d. Especially, the sensor with PT for 30 s demonstrated the best performance in durability over the entire range of the cyclic test (Figure 8b).

Sensing Mechanism
To understand how our multi-layer PPy/PDMS sensors respond to an external load of stretching, the morphological change on the surface of PPy films (fabricated with O2 PT for 30 s) under different levels of elongation (ε = 0, 10, 20, 30, 40, and 50%) were more closely observed using optical microscopy at a ×100 magnification as shown in Figure 9. Under no elongation, PPy multi-layer films obviously show no microcrack and sit stable on the PDMS substrate. When the sensors are stretched to a strain level of 10%, narrow line-shaped microcracks (aligned to the perpendicular direction with elongation) start to appear on the PPy film surface as shown in Figure 9b. Average microcrack line width (wc) observed at this strain was 2.5 μm, where wc was measured from four different areas (50 × 50 μm 2 ) of each sensor structure. With the increase of strain level, the wc was increased significantly as well as the population of microcracks, and measured wc at 20, 30, 40, and 50% was 8.0, 15.6, 18.0, and 20.0 μm, respectively (see Figure 9c,d).

Sensing Mechanism
To understand how our multi-layer PPy/PDMS sensors respond to an external load of stretching, the morphological change on the surface of PPy films (fabricated with O 2 PT for 30 s) under different levels of elongation (ε = 0, 10, 20, 30, 40, and 50%) were more closely observed using optical microscopy at a ×100 magnification as shown in Figure 9. Under no elongation, PPy multi-layer films obviously show no microcrack and sit stable on the PDMS substrate. When the sensors are stretched to a strain level of 10%, narrow line-shaped microcracks (aligned to the perpendicular direction with elongation) start to appear on the PPy film surface as shown in Figure 9b. Average microcrack line width (w c ) observed at this strain was 2.5 µm, where w c was measured from four different areas (50 × 50 µm 2 ) of each sensor structure. With the increase of strain level, the w c was increased significantly as well as the population of microcracks, and measured w c at 20, 30, 40, and 50% was 8.0, 15.6, 18.0, and 20.0 µm, respectively (see Figure 9c,d).
cially by breaking the interfacial adhesion of PPy thin film and random fracturing in the parts of the PPy networked fiber structure; therefore, this process could achieve the effective release of stored strain energy. In the release state, as the PPy/PDMS elastic body is contracted again, the line width of the microcracks generated in the PPy tissue was reduced and even closed. This gave rise to the recovery of many "current paths" and considerable reduction in electrical resistance as illustrated in Figure 9g-l.

Human Motion Monitoring
Due to the electromechanical sensing performance and reliability as discussed above, we were able to examine the feasibility of applying the multi-layer PPy/PDMS sensors fabricated with O2 PT to the monitoring of human body movements. Figure 10 demonstrates the change in the resistance signal of the sensor generated by various movements of human-body joints. The strain sensor was first attached to the upper part of an index finger (see Figure 10a), and the corresponding change in normalized resistance was measured according to the finger's bending movements (bending angle = 10, 30, 45, 60, and 90°). The resistance signals of the sensor exhibited a significant step-like increase as the bending strain changed toward the final bending by 90°. When the finger was fully extended, the resistance immediately returned to its original level. The greater the degree of bending, the higher the relative resistance changed, showing high strain sensitivity (GF of few hundreds) and ultrafast response capability (in few tens of milliseconds). In addition, our PPy/PDMS sensors could effectively monitor subtle human motions such as drinking, swallowing, and chewing. As an example, Figure 10b displays the real time response of The formation of microcracks in PPy film structure under a strained state is caused by the random dissipation of strain energy applied to the porous PDMS elastomers, especially by breaking the interfacial adhesion of PPy thin film and random fracturing in the parts of the PPy networked fiber structure; therefore, this process could achieve the effective release of stored strain energy. In the release state, as the PPy/PDMS elastic body is contracted again, the line width of the microcracks generated in the PPy tissue was reduced and even closed. This gave rise to the recovery of many "current paths" and considerable reduction in electrical resistance as illustrated in Figure 9g-l.

Human Motion Monitoring
Due to the electromechanical sensing performance and reliability as discussed above, we were able to examine the feasibility of applying the multi-layer PPy/PDMS sensors fabricated with O 2 PT to the monitoring of human body movements. Figure 10 demonstrates the change in the resistance signal of the sensor generated by various movements of human-body joints. The strain sensor was first attached to the upper part of an index finger (see Figure 10a), and the corresponding change in normalized resistance was measured according to the finger's bending movements (bending angle = 10, 30, 45, 60, and 90 • ). The resistance signals of the sensor exhibited a significant step-like increase as the bending strain changed toward the final bending by 90 • . When the finger was fully extended, the resistance immediately returned to its original level. The greater the degree of bending, the higher the relative resistance changed, showing high strain sensitivity (GF of few hundreds) and ultrafast response capability (in few tens of milliseconds). In addition, our PPy/PDMS sensors could effectively monitor subtle human motions such as drinking, swallowing, and chewing. As an example, Figure 10b displays the real time response of resistance upon the very small-scale motion of water drinking when collected from the sensor mounted around a human throat as shown in the inset. resistance upon the very small-scale motion of water drinking when collected from the sensor mounted around a human throat as shown in the inset.

Conclusions
In conclusion, a chemical oxidative polymerization synthesis method combined with three-step O2 PT was demonstrated in this study for the fabrication of highly sensitive and reliable multi-layer PPy/PDMS strain sensor structures. The sensors fabricated with O2 PT under an optimized condition showed high sensitivity (GF = 438), a wide sensing range up to 50% elongation, good electrical conductance of PPy films in our sensing strain range of 0-50% (R < 300 kΩ), fast response time (~50 ms), reliable sensing stability at different strain rates, and long-lasting sensing durability upon the cyclic stitching/release. This sensor performance improvement is due to the oxygen plasma exposure to the growing PPy films, which successfully reinforce the bonding characteristics of PPy with underlying layers and provide stronger resistance force to microcrack generation under strained state. Successful extension of this sensor fabrication technology will be applicable to future realtime monitoring of human bodies, flexible robots, and wearable electronic devices and very large and diverse movements.

Acknowledgments:
The authors would like to thank the former collaborating students for their invaluable help and ideas.

Conclusions
In conclusion, a chemical oxidative polymerization synthesis method combined with three-step O 2 PT was demonstrated in this study for the fabrication of highly sensitive and reliable multi-layer PPy/PDMS strain sensor structures. The sensors fabricated with O 2 PT under an optimized condition showed high sensitivity (GF = 438), a wide sensing range up to 50% elongation, good electrical conductance of PPy films in our sensing strain range of 0-50% (R < 300 kΩ), fast response time (~50 ms), reliable sensing stability at different strain rates, and long-lasting sensing durability upon the cyclic stitching/release. This sensor performance improvement is due to the oxygen plasma exposure to the growing PPy films, which successfully reinforce the bonding characteristics of PPy with underlying layers and provide stronger resistance force to microcrack generation under strained state. Successful extension of this sensor fabrication technology will be applicable to future realtime monitoring of human bodies, flexible robots, and wearable electronic devices and very large and diverse movements.