Fig. 2 shows SEM images of the morphology of the AgNWs-coated fabric with different magnification. As shown in Fig. 2a, the yarns of the cotton were layered naturally with porous structure. The outmost surface of the fabric is covered by AgNWs (Fig. 2b), on which nanowires are uniformly wrapped on the fibers. Particularly, between the neighboring yarns, there are empty spacing that are bridged by the attached AgNW conductive networks (Fig. 2c). To be noticed, long and uniform wires were observed between adjacent yarns and the average diameter of AgNW is around 55 nm. In Fig. 2d, the AgNWs are homogenously formed at main area of the yarn surface, while disconnected at some point due to the poor adhesion. Furthermore, the distant between nanowires adhered on the individual yarn is relatively larger than that of the nanowires between neighboring yarns.
Also, the density of the AgNWs on the surface of the fabrics was adjusted by the times of dip-coating cycles. The AgNWs-coated fabric with 1 dip cycle and 5 dip cycles were shown in Fig. S1 and Fig. 2d, respectively. Compared with the high density one, the nanowire mesh spacings of 1 dip cycle were increased from below 1 µm to 2-4 µm.
The composition of the AgNWs-coated fabric was also investigated by the EDS, as illustrated in the inset of Fig. S2. In addition to the C and O contents which mainly attributed from cotton, Ag element was also observed, indicating the distribution of AgNWs on the cotton.
The sensing principle of the pressure sensor is shown in Fig. 3a and the cross-section SEM images of the sensor with different pressures are shown in Fig. 3b-e. In the unloading state, the initial resistance is large, which is caused by the non-contact AgNWs on the fabrics (Fig. 3b). Once the pressure was applied, the increasing fiber-scale contacts of nanowires on the adjacent fabrics contributed to a decrease in the resistance (Fig. 3c). Furthermore, when the nanowires on the fabrics were contacted completely, the continued pressure loaded on the fibers then increased the yarn-scale contacts. As Fig. 3d shown, the yarns length in Y direction are reduced from about 200 µm to 160 µm, confirming the compression between the yarns. According to the cross-section SEM images, the AgNWs were formed both on the yarn surface and inside the adjacent yarns (Fig. S3). When the pressure was applied and the yarns were compressed, the AgNWs inside the yarns could contact and further reduced the resistance of the sensor. With loading pressure increasing, the neighboring fabrics were compressed (Fig. 3e), fabric-scale contacts further reduced the resistance of the sensor due to the increased contact area between the facing fabrics. At this point, the total thickness of the double-layered fabrics was reduced from 600 µm to 350 µm. Therefore, the pressure sensing of the sensors was determined by the synergistic effect of the fiber/yarn/fabric multi-scale contact. These cross-section SEM images further confirmed the pressure sensing mechanism.
The influence of bending on the surface morphology of the AgNWs coated fabrics was investigated by cross-section SEM images shown in Fig. S4. With the little bending deformation, there is no obvious crack and peeling-off problem of AgNWs network on the fabrics (Fig. S4b) compared with initial state (Fig. S4a). In order to further investigate the influence of bending deformation, the SEM images of AgNWs-coated fabrics with 500 times bending cycles was taken and shown in Fig. S5. Fig S5 shows many delaminated spots which potentially occur device degradation. This result indicates that the stability of the AgNWs coated fabric need to be further improved in the future.
Fig. 4a shows the current-voltage curves of the pressure sensor under different pressures. When the applied pressure increased from 0 kPa to 100 kPa, the resistance of the sensor was decreased. Furthermore, the response of sensor was steady and fell in line under Ohm’s law . The current of pressure sensor is shown in Fig. 4b, which is relatively constant under different applied pressure, revealing that the response of the sensor is stable for different pressures. Therefore, the results provide excellent electrical stability for the potential application of the pressure sensor.
To investigate the performance of pressure sensors, the relative current changes (ΔI/I0) versus pressure with the different AgNWs dip-coating cycles and mesh hole diameters were shown in Fig. 4c and d. Here, the sensitivity of the pressure sensor was defined as S = (ΔI/I0)/P, where P denotes the applied pressure. At a mesh hole diameter of 0.25 mm, the sensitivity of the pressure sensor was strongly dependent on the AgNWs dip-coating cycles. The sensitivity of the sensors was improved from 2.12×103 kPa-1 to 1.98×105 kPa-1 within the range of 0–10 kPa when the dip-coating cycles increased from 1 to 5. In addition, the sensitivity improved from 764 to 1.12×103 kPa‒1 at 10–100 kPa. The improvement of sensitivity with high dip-coating cycles is mainly attributed to the increase of the AgNWs densities.
Furthermore, the dependence of hole diameter was subsequently characterized. The pressure sensors with 5 dip-coating cycles exhibited enhancing sensitivities with increasing diameters, which were increased from 1.12×103, 9.88×103, to 2.16×104 kPa‒1 within the pressure range of 10–100 kPa, respectively. The enhancement in the sensitivity was mainly attributed to the increased contact area through the larger holes. However, once the diameter exceeded 1 mm with 4 cycles, the initial interface of the facing fabrics resulted in more contact in the unloading state, thus significantly lowering the contact resistance between the fabrics (Fig. S6). Furthermore, when the thickness of the spacer cotton is changed, the performances of the sensors get worse (Fig. S7). The sensor with lower thickness shows a decrease of ∆I/I0 due to the contacting of the facing fabrics in the initial state (Fig. S7a). In addition, higher thickness reduces the contact of the fabrics. When the thickness of spacer cotton increased to 1 mm, the AgNWs on the fabrics were not contact until the pressure exceed 10 kPa (Fig. S7c).
Notably, the pressure sensor clearly exhibited two linear current segments, the sensitivity increases sharply in the low-pressure range and increases gradually in the high-pressure range. According to the sensing principle as we mentioned above, in the low-pressure range, the adjacent AgNWs contact plays an important role in increased current. Furthermore, when the pressure is increased to 10-15 kPa, the AgNWs on the interface were contact completely. The current changes were mainly determined by the contact resistance between the yarns and fabrics, which was relatively stable. Contact between the yarn- and fabric-scales played a more significant role in the sensing mechanism at a dip-coating cycles of 5 times and diameter of 0.75 mm by enhancing the sensitivity and linear range. Therefore, the diameter of 0.75 mm (Fig. 4d) exhibited a higher sensitivity and larger linear range due to increased contact [33, 34].
The dynamic response of the device was studied under loading/unloading pressure cycles. The sensor exhibited an immediate response to the cyclic pressures. The time-resolved response was analyzed to quantify the response and relaxation times (Fig. 5a). The measured response and relaxation times were 32 and 24 ms, respectively. The performance of the sensor under difference pressure is also investigated and shown in Fig. S8. The sensor clearly distinguished a subtle pressure of 50 Pa, indicating the excellent performance of the sensor. The ΔI/I0 with an applied pressure of 10 kPa with 1,000 loading cycles was used to verify the repeatability of the device (Fig. 5b). The results show the excellent stability of the pressure sensors. Furthermore, the air permeability of both normal cotton and AgNW coated cotton was investigated. Despite the air permeability was reduced from 787.3 mm/s to 252.6 mm/s, this value is still much higher than the recently reported ones [35,36]. This result demonstrated that the wearable pressure sensors based on silver nanowire-coated fabrics remains good air permeability because of its high porosity.
Due to the natural flexibility of fabrics and high sensitivity of sensors, the pressure sensor was wearable and able to detect mechanical signals such as physiological pulse and respiratory rate. First, the device was attached at the wrist using an adhesive bandage to monitor the pulse pressure. Fig. 5c presents the real-time recorded data, in which the pulse rates were measured to be ≈72 beats min‒1. In addition, the sensor was also attached to a mask to detect respiration states. Fig. 5d indicates that the normal respiratory rate of 10 breaths per minute of an adult and a square-like wave for normal breathing. Furthermore, the width of the waveband indicated the maintained time of breathing. These results suggest that the pressure sensor with high sensitivity and superiority has great potential in wearable healthcare device applications.