PEDOT:PSS/Natural Rubber Latex Inks for Screen‐Printing Multifunctional Wearable Strain–Humidity Sensors

Wearable strain–humidity sensors are increasingly recognized for their role in healthcare and detecting human motion signals. However, many strain–humidity sensors require complex manufacturing processes and have limited applicability. In this study, based on the good miscibility between natural rubber latex (NRL) and poly (3,4‐ethylenedioxythiophene):poly (4‐styrenesulfonate) (PEDOT:PSS), PEDOT:PSS/NRL composite conductors are transferred onto fabric substrate through screen printing and prepared high‐performance strain–humidity sensor. The PEDOT:PSS/NRL blends strain–humidity sensor exhibited good stability and durability (500 cycles) at 10% strain, achieves a gauge factor (GF) of 123.8 with excellent linearity, making it operational in detecting various human motions, such as finger bending, elbow bending, wrist bending, and so on. In addition, the PEDOT:PSS and water molecules interaction results in a humidity response time as low as 0.72 s and recovery time as low as 0.85 s for the PEDOT:PSS/NRL composite strain–humidity sensor, and a wide relative humidity detection range (6–83%), allowing it to be utilized for precise detection of human breathing. Furthermore, it can monitor the use's joint movement, facial muscle movement, and respiratory rate, which show its potential application prospect in health monitoring.

(PAA), [23] polyethylene glycol (PEG), [24] or flexible conductive material, such as silver nanowires (Ag NWs) [25] into PEDOT: PSS.For example, Xu et al. [26] reported a stretchable sensor comprising of carbon nanocoils (CNCs) and PEDOT:PSS/WPU.The sensor exhibits a high sensitivity with GF of 25 over a large strain range of 0-50%.Gao et al. [27] prepared PEDOT: PSS/PVA composite conductive fibers through the wet-spinning method, with a maximum tensile strength of %180 MPa and a maximum elongation at break of %32%.Liang et al. [26] reported on a strain sensor incorporating PEDOT: PSS, MXene (Ti 3 C 2 T X ), and PAA exhibiting stretchability up to 1000%, and a GF value of 20.86.Li et al. [28] prepared PEDOT:PSS/PEG films using spin-coating method, with a strain of up to 45%.In addition, Shen et al. [29] described a strain sensor based on PEDOT:PSS, Ag NWs, and MXene nanosheets with a stretchability of up to 120% and GF value of 37.44, which further improved conductivity.Similarly, Thana et al. [30] reported a strain sensor based on PEDOT:PSS and Ag NWs, showing a sensitivity of 418 and a strain range of 50%.However, most polymers and additives suffer low adhesion to substrates and limited resilience, leading to rupture and failure under stretching.Alternately, natural rubber that produced from NRL has significant strength, excellent resilience, and considerable stretching, [31] and it has already been studied as a natural matrix for nanocomposites. [32]As a low-cost, environmentally friendly base material, nature rubber has good elasticity and can quickly return to its original shape after releasing strain. [33]For example, Boratto et al. [34] reported a highly conductive and stretchable polymer by blending PEDOT:PSS with NRL, which can withstand 700% strain.Yang et al. [13] fabricated a stretchable conducting elastomer composed of PEDOT:PSS and NRL, with a maximum electrical conductivity of 87 S cm À1 and a maximum strain capability of 490%.More importantly, it can provide a strain sensor that is very suitable for human use and allows attachment to different parts of the human body.Based on this, NRL can be used as an excellent flexible substrate for wearable sensing applications.[37] Significant progress has been made in previous studies on elastic sensors, for example, Foroughi et al. [38] prepared fabric strain sensors using a continuous wrapping method.The resulting fabric exhibited a strain sensing range of up to 100% and a gauge factor (GF) of approximately 0.4.After 1000 stretching-releasing cycles at 100% strain, the sensitivity decreased by less than 2.3%.Li et al. [39] reported the successful fabrication of a textile sensor using plating and spray-coating techniques that can detect human respiratory rate in real time.However, the functionality of the sensors in the aforementioned examples is limited to the measurement of a single signal.In practical applications, to meet the demands of multiple human scenarios, various signals need to be collected for a comprehensive assessment of the patient's health status, for instance, joint bending, facial muscle movement, and breathing frequency are among the signals that need to be monitored.However, the research in this area is still in its early stages.
Here, we adopted a simple screen printing method to prepare a strain-humidity multifunctional wearable sensor consisting of PEDOT:PSS/NRL.The print adaptability of PEDOT:PSS/NRL ink was studied by rheological tests.Then, the ink was directly printed on NRL-treated fabric to prepare the PEDOT:PSS/NRL strain-humidity wearable sensor.The mechanical and electrical performance of the strain-humidity sensor under different tensile strain conditions was further studied.PEDOT:PSS was used as a functional material to detect environmental humidity.Due to the formation of hydrogen bonds between PEDOT:PSS and water molecules, [40] the sensor exhibits characteristics such as a wide humidity range, fast response speed, and good robustness to humidity.The PEDOT:PSS/NRL strain-humidity sensor is sensitive to tensile and humidity changes, and there are different humidity change trends under different tensile strains.By comparing tensile and humidity changes, the applied stimuli can be well distinguished.

Result and Discussion
As shown in Figure 1a, the composite conducting ink was prepared by simply mixing PEDOT:PSS and NRL dispersions.PEDOT:PSS and NRL can be mixed in a large weight percentage of PEDOT:PSS, which is up to 30 wt%.In those composite inks, NRL maintained the consistent nanosphere size and their diameters matched their dynamic light scattering results before mixing (Figure S2a, Supporting Information).By studying the zeta potential of PEDOT:PSS/NRL composite inks, we found that PEDOT:PSS (À42.6 mV) has a negatively charged surface and it repelled the negatively charged NRL (À38.3 mV), a zeta potential of À49.3 mV for the PEDOT:PSS/NRL ink, confirming the stability of the inks (Figure S2b, Supporting Information).
To understand the interaction and component ration of PEDOT:PSS/NRL composite conductors, Fourier transform infrared (FTIR) spectroscopy and thermal gravimetric analysis (TGA) were carried out.As illustrated in Figure S3a, Supporting Information, the FTIR spectrum of NRL showed two typical peaks at 841 and 2962 cm À1 corresponding to =CH 2 and -CH 3 , respectively. [41]PEDOT:PSS exhibited peaks belonging to the stretching vibration of C-O-C (at 1063 cm À1 ) and the symmetrical vibration of sulfonic acid group in PSS (at 1162 cm À1 ). [42,43]Obviously, the mixed nanocomposite combines the characteristics of both PEDOT:PSS and NRL.The thermal stability of the PEDOT:PSS/NRL composite was measured using TGA (Figure S3b, Supporting Information), and the weight loss of PEDOT:PSS can be divided into three distinct regions.In the first region from 40 to 150 °C, the weight loss is attributed to the evaporation of water in PEDOT:PSS/NRL composites. [44]The second weight loss observed between 230 and 370 °C is caused by the degradation of the sulfonate group in PSS decomposition. [45]The third region between 380 and 600 °C is related to the decomposition of the polymer backbone of PEDOT:PSS. [46]In contrast, NRL exhibits a significant weight loss phenomenon at 400 °C. [47]By measuring the weight percentage at around 450 °C, we confirmed the weight ration of PEDOT:PSS and NRL in the composite matches the desired ratio, which also proven the uniform mixing of the composite ink.
To determine the optimal ratio of PEDOT:PSS content, we tested the electrical conductivity of composite inks at increasing PEDOT:PSS loading.The conductivity increased from 2.4 to 10 S cm À1 when the PEDOT:PSS loading reached 50 wt%.Afterward, the conductivity significantly increased to 159 S cm À1 at PEDOT:PSS loading of 90 wt%, suggesting the formation of conducting network in the mixture.However, we found that when PEDOT: PSS loading reached 50 wt%, the dried composite ink showed cracks at elongation of less than 10% (Figure 1f ).Therefore, this work will control the PEDOT:PSS load below 50 wt%.We further compared the composites with 30 and 10 wt% of PEDOT: PSS.At 10 wt%, the resistance shows a grate fluctuation without clear trend as the strain increases (Figure S4a, Supporting Information), and the resistance of the tensile hysteresis loop at 10% of strain shows irregular and unstable changes (Figure S4b, Supporting Information).This may be caused by the poor connection between PEDOT and PSS in the mixture, resulting in an uneven distribution of the internal conductive network in the sample and unstable resistance changes.At the PEDOT:PSS loading of 30 wt%, the resistance shows a consistent increasing trend with increasing strain (Figure S4c, Supporting Information) and the structure remains intact under 10% strain (Figure 1e).Therefore, the PEDOT:PSS loading of 30 wt% was chosen as the optimal proportion for conducting inks and sensing applications.
To evaluate the printing adaptability of the PEDOT:PSS/NRL ink at PEDOT:PSS loading of 30 wt%, we measured the viscosity using a rheometer.Figure 2a shows that the viscosity curve of the ink exhibits shear-thinning behavior, and its viscosity decreases with increasing shear rate, indicating that the ink was a non-Newtonian fluid and was suitable for screen printing. [48]dditionally, in screen printing, the ink should have the ability to gradually restore its viscosity after the removal of stress or shearing force. [49]Therefore, we studied the thixotropic behavior of the ink in the peak holding step test with different periodic shearing rates.It can be observed that as the shear rate was applied, the viscosity of the ink quickly dropped, and it recovered immediately after the share rate was removed.And this trend applies to a wide range of share rates from 200 to 900 s À1 (Figure 2b).The ink was further characterized by oscillatory rheology measurements using a stress sweep step test (Figure 2c).The curves can be divided into two regions. [50]In Region I, the storage modulus (G´) and loss modulus (G 00 ) of the ink was clearly separated at shear stresses below 96 Pa, indicating the formation of a slender network structure dominated by elasticity in the ink.Although the ink exhibits elasticity dominated behavior in Region I, starting from Region II, as the shear force increases, G 00 becomes higher than G´, indicating that the liquid behavior of ink dominates. [51]In the frequency range of 0.1-100 Hz, the diagram shows G 00 of the ink was invariably greater than G 0 , which illustrates that the ink maintained its viscosity characteristic for the selected frequency range (Figure 2d).The results of rheological analysis show that PEDOT:PSS/NRL ink at 30 wt% loading of PEDOT:PSS have good printability.By simply changing the printing patterns, the PEDOT:PSS/NRL ink can be easily printed on fabric for constructing flexible sensors with different patterns and shapes to meet different needs, as shown in Figure 2e.
To investigate the tensile properties of fabric substrate, we conducted a series of mechanical performance tests.The maximum tensile strength of the original elastic fabric was 7.4 N, while the maximum tensile strength after NRL treatment was 11.2 N (Figure 3a).The mechanical properties of the fabric were significantly improved after the NRL treatment because natural rubber increases the interaction between filaments in fabric. [52]fter the processing of NRL, the maximum elongation of the fabric is 55%.Therefore, we investigated the macroscopic mechanical properties degradation of the fabric during cyclic stretching and unloading process with a deformation of 50%. Figure 3b shows the strain-stress curve of the fabric substrate under one complete cycle with 10-50% strain.As the tensile stress increases, the elastic modulus of the fabric substrate does not decrease significantly, indicating that the fabric substrate exhibits good elastic recovery performance.For long-term use, the fabric must exhibit both mechanical properties under single stretch and cyclic mechanical properties under maximum deformation.In Figure 3c, the prepared fabric substrate's tensile behavior stabilizes in the second stretching cycle, and the maximum stress in consecutive cycles showed a minimum decrease, indicating a stable substrate structure during tensile tests.
Due to the knitting pattern of the fabrics is anisotropic, [53] mechanical testing was performed on samples taken from both the course and wale directions.Tensile force was consistently applied in a single direction, as shown in Figure S5, Supporting Information, when the fabric was strained along the wale direction, the strain reached its maximum value of 50%, which is higher than that of course direction.Also, at wale direction, the fabric shows higher maximum stress of 0.96 MPa.This indicates that the fabric is more resistant to stretching in the wale direction than in the course direction.Therefore, we designed the strain sensor to measure strain in the wale direction while the humidity sensor measures humidity in the course direction to minimize the strain range.
Subsequently, the characterizations of PEDOT:PSS/NRL strain sensing are carried out.Typically, the sensitivity of PEDOT:PSS/NRL composite conductors strain sensor is evaluated in terms of the GF value.The GF value is defined as the ratio of the relative change in resistance to the applied strain, and it is calculated using Equation ( 1) where ΔR = RÀR 0 , R 0 and R represent the initial resistance and resistance after deformation of the PEDOT:PSS/NRL composite conductors, and ε is the applied strain. [54]Figure 3d displays variation trend of the strain sensor resistance at a strain of 50% and strain speed of 10 mm min À1 .A linear increase of ΔR/R 0 is defined between 0% and 50% strain (ΔR/R 0 = 123.8Â ε À 0.01; the degree of fitting is 0.974), corresponding to a GF of 123.8.
In addition, the study systematically investigated the electric cyclic response of the strain sensor under various dynamic strains (Figure 3e).The highly consistent signal response pattern under various strains indicates that the strain sensor possesses excellent stability and continuous response capability to the multiple cyclic loading.The cyclicity and durability of the strain sensor were further investigated.Figure 3f illustrates that the PEDOT:PSS/NRL strain sensor with a loading-unloading process of 10% strain for 500 cycles, at a strain rate of 10 mm min À1 .The output resistance signal demonstrated a stable linear variation, indicating outstanding cycling stability of the sensor.This suggests favorable reproducibility and robustness of the sensors in practical applications.Notably, as the strain increased to 50%, the ΔR/R 0 value decreased, which could be attributed to the destruction and reconstruction of the conductive networks in sensor (Figure S6a, Supporting Information).Additionally, the strain sensor exhibited satisfactory stability at a 50% strain under slower strain rates of 0.5, 1, and 3 mm min À1 (Figure S6b-d, Supporting Information), suggesting its suitability for detecting signals at various deformation rates.
Strain sensors are vital medical monitoring tools that can detect a wide range of bodily movements with consistency. [55] study was conducted to investigate the applications of the strain sensor, which has high sensitivity and stable cyclic response, in recognizing human joint motion and facial expressions.Figure 4a shows a significant increase in relative resistance when the finger is bent at a 90°angle compared to a 45°angle, as demonstrated by the strain sensor.This sensor also exhibits varying response strength with different angles of finger bending and returns to its initial resistance value when the finger is straightened.In addition to finger movements, the strain sensor can monitor joint movements of the wrist and elbow (Figure 4b,   c).Additionally, the strain sensor can be affixed to the skin near the mouth to detect various facial expressions, such as chewing, puffing cheeks, and transitioning from a poker face to a smiling face (Figure 4d-f ).These results demonstrate that the PEDOT:PSS/NRL composite conductors strain sensor, which was fabricated by screen printing, has tremendous potential for use in facial motion detection.
Due to the wettability of PEDOT: PSS in the composite ink, the printed inks are capable to be used as humidity sensor.To determine the response range of the sensor, we conducted humidity-sensitive experiments at room temperature using a self-made static humidity sensing device (Figure S7, Supporting Information).Due to the hydrophilic groups present in PEDOT:PSS, [56] the PEDOT:PSS/NRL humidity sensor exhibited a sensitive humidity response over a wide range, from 6% to 83% of RH (Figure 5a).Furthermore, we also noted that the resistance of the humidity sensor decreased as the relative humidity increased.This behavior can be attributed to the enhancement of charge-carrier conduction within the polymer.The presence of water molecules on the surface may lead to an increase in the density of charge carriers, which can be explained by the relatively high dipole moment of water molecules. [57]As shown in Figure 5b, the humidity sensor exhibits no significant fluctuations or changes in resistance values when they exposed each humidity condition, indicating that the sensor is capable of detecting a constant humidity level.To determine the response and recovery speeds of the PEDOT:PSS/NRL composite conductors humidity sensor, we inferred the response time (t res ) and the recovery time (t rec ) from the sensing curve.We found that the PEDOT:PSS/NRL humidity sensor exhibits an extremely fast response time (<0.72 s) and recovery time (<0.85 s) when RH increases from 23% to 74% (Figure 5c).When the sensor is exposed to an environment with low RH (e.g., 23%), the resistance value is 3.39 MΩ.However, as the RH increases to 74%, the resistance decreased to 0.82 MΩ.It can be explained  that high humidity leads to the expansion of the sensor volume, thus reducing the distance between PEDOT:PSS chains, promoting the movement of charge carriers, and reducing the electrical resistance. [58]In contrast, when RH returns to 23%, the PEDOT: PSS layer undergoes desorption, and its original volume is restored, resulting in a quick recovery of electrical resistance.Both the response time and the recovery time are approximately equal indicating negligible negative hysteresis.To differentiate the sensor response signals to strain and humidity, we studied the pattern of sensor response to humidity under strain condition of 10%, 30%, and 50%.It was evident that the sensor exhibited corresponding changes in humidity response under different stretching conditions (Figure S8a, Supporting Information).Furthermore, the sensor exhibits stability in humidity response under different strains (Figure S8b, Supporting Information).
Interestingly, the sensor response to RH values was higher at a strain of 50%, which may be due to the fact that stretching the sensor can expose a larger specific surface area, providing more adsorption sites for water molecules.Additionally, we tested the ability of the humidity sensor to detect human breath.The humidity sensor is capable of distinguish respiratory rate and depth during slow, normal, and rapid breathing states. [59]The PEDOT:PSS/NRL humidity sensor has several advantages such as its small size, flexibility, and lightness.It can be attached to a mask, allowing for convenient monitoring of human breath.Figure 5d-f represent the three breathing states, where the first part (light pink) of the entire breathing process corresponds to slow breathing with an approximately interval of 7.5 s, the second part (light green) corresponds to normal breathing with an approximate interval of 3.9 s, and the third part (light orange) corresponds to breathing intervals of approximately 1.5 s.It can be observed from the curves that the impedance change in the third part is the smallest, with a shorter interval time than the other parts have, which is consistent with the actual situation.This proves that the sensor has a great potential in timely monitoring of respiratory rate and depth. [60]

Conclusion
In summary, we have developed a wearable sensor using screen printing technology that can detect both strain and humidity.We prepared a water-based PEDOT:PSS/NRL conductive ink that demonstrated excellent printability.We conducted a characterization of the mechanical and humidity-sensitive properties of the sensor.Our findings revealed remarkable stability and durability, as evidenced by the ability of the sensor to endure 500 cycles at a 10% strain, while also achieving a high GF of 123.8.Furthermore, the sensor exhibited fast humidity response times, as low as 0.72 s, and recovery times as low as 0.85 s, with a wide relative humidity detection range of 6-83%.Notably, the sensor could also monitor joint movements, facial muscle movements, and respiratory rates, indicating its promising potential in health monitoring applications.
Preparation of the PEDOT:PSS/NRL Composite Inks: Briefly, different concentrations of PEDOT:PSS were mixed with NRL to prepare the composite inks.For instance, 5 vol% DMSO was added into aqueous PEDOT: PSS (1.3 wt%) and stirred for 1 h.Afterward, the pH of the PEDOT:PSS solution was adjusted to 8 by adding NH 3 •H 2 O.The aqueous NRL (60 wt%, 3 mL) was magnetically stirred at room temperature for 1 h before it was added to the PEDOT: PSS solution drop wisely with mild stirring for 0.5 h and subjected to ultrasound for 1 min.By controlling the amount of NRL, we achieved composite inks with different PEDOT:PSS ratios.The obtained viscous composite inks were used for further experiments.
Screen Printing of Fabric-Based Sensors: First, the fabric was washed with deionized (DI) water and 0.1 M NaOH several times to remove the oil or glues that may be contaminated during fabric production.Afterward, the fabric was rinsed with DI water and subsequently air-dried in an open environment.To prevent infiltration of PEDOT:PSS/NRL composite inks into the fabric, a diluted NRL (10 wt%) was used to rinse the fabric and form a thin layer of rubber.The screen-printing process and the PEDOT:PSS/NRL ink was printed using a 10 Â 10 in.precision screen plate (250 mesh counts) with various patterns.After printing, the screen plate was removed and the sensor pattern was dried by placing the fabric in an oven at 120 °C for 10 min to remove extra water and vulcanize the NRL.
Characterization: A bench top digital multimeter (Keysight 34465A) and a two-probe method was used to measure the resistance of the specimen and calculate its conductivity according to the sample size.Field-emission scanning electron microscopy (SEM) of the sensor was performed on a Hitachi S-4800.Element mass percent values were analyzed with a Vario EL Elementar cube (Elementar Analysensysteme GmbH, Germany).Zeta potential of PEDOT:PSS and NRL was measured using a Malvern Zetasizer Nano ZS90.Malvern Zetasizer Nano ZS90 was also used to measure the size distribution of PEDOT and latex particle.The TGA data were recorded using a TG STA 449C thermal analyzer (Netzsch, Germany) with a heating rate of 10 °C min À1 from 25 to 700 °C.FTIR spectroscopy was conducted in the range of 300-4000 cm À1 using Nicolet Nexus 870 (Thermo Fisher Scientific).The rheological properties of the prepared inks were measured by a HAAKE MARS60 rotary rheometer.The mechanical properties of the sensor were measured using a universal tensile testing system (Tensile Tester TM2101-T5) with controlled strain range and frequency.All the device performance measurements on human body were carried out following the ethical policy and we acquired the oral approval from all participants to collect the data for publication.

Figure 2 .
Figure 2. a) Viscosity as a function of shear rate for PEDOT:PSS/NRL ink.b) Thixotropic behavior of the ink during the screen-printing process where the viscosity drops and recovers after different high shear rate (200, 300, 500, 700, and 900 s À1 ) are clearly evident.c) Variation of G 0 and G 00 with shear stress for the ink.d) The variation of G 0 and G 00 of PEDOT:PSS/NRL ink with oscillation frequency.e) Photograph of printed patterned PEDOT: PSS/NRL ink patterns.

Figure 3 .
Figure3.a) Stress-strain curve of fabric before and after being treated with natural rubber latex (NRL).b) Loading-unloading measurement of the fabric under varied strains (only one cycle) without a resting interval; and c) 50 successive cyclic loading-unloading curves at a tensile strain of 50% without a resting interval between two consecutive tests.d) ΔR/R 0 upon the applied strain for PEDOT:PSS/NRL composite conductors.e) Multicycle tests of ΔR/R 0 of the conductor upon stretching to different maximum strains from 1% to 50%.f ) Long-term stability of the conductor at the PEDOT:PSS loading of 30 wt% under 10% stretching cycles at rate of 10 mm min À1 for 500 times.

Figure 4 .
Figure 4. Response patterns of the PEDOT: PSS/NRL composite conductors strain sensor fixed on a) finger, b) wrist, and c) elbow under diverse bending degrees.Response patterns of the PEDOT:PSS/NRL composite conductors strain sensor when the volunteer d) chew, e) blow, and f ) smiles.

Figure 5 .
Figure 5. a) The resistance of PEDOT:PSS/NRL composite conductors humidity sensor versus relative humidity curves.b) Stability of humidity sensing upon exposure to different levels of relative humidity.c) Response and recovery time of the humidity sensor when relative humidity increases from 23% to 74%.The response and recovery curve at d) slow, e) normal, and f ) fast breathing rate.