Page 118 - Read Online

Page 8 of 13 Zhang et al. Soft Sci. 2025, 5, 17 https://dx.doi.org/10.20517/ss.2024.68

In practice, the sensitivity (S) is measured as S = (ΔI/I )/ΔP where ΔI is the relative current change with
0
pressure, I is the current at no pressure, and ΔP is the applied pressure. Figure 3B shows the I-V curves with
0
good linear characteristics over up to 100 kPa, indicating a stable piezoresistive response and the ability to
discriminate different pressures over a wide detection range. Figure 3C plots the sensitivity curve across the
whole detection range, which mainly consists of three linear regions: 16.7 kPa within 0-20 kPa, 10.6 kPa
-1
-1
-1
within 20-40 kPa, and 6.1 kPa within 40-100 kPa. The explanation can be described as follows: in the first
linear region, without compression, the porous space between paper fibers is very large. Upon initial
compression, the deformation of the porous network is prominent, starting to get MXene connected with
formation of many conductive contacting points. Therefore, a small force can induce a large resistance
reduction, leading to a large sensitivity. In the second linear region, as compression continues, the
conducting pathways formed by MXene connection tend to saturate. Further compression would cause the
porous network to further deform with paper fibers to contact each other. The network tends to become
compact and the resistance reduces at a slower rate, leading to a medium sensitivity. In the third linear
region, after all the voids are closed, further compression would shorten the thickness of the compact
MXene/paper fiber solid sheet. The resistance can be further reduced but at a very slow rate, leading to a
small sensitivity. The influence of MXene dip-coating cycles on the sensitivity was also evaluated. As the
dip-coating time increases, the sensitivity correspondingly increases [Supplementary Figure 5], because
decoration of more MXene leads to more conductive pathing ways in the network. Figure 3D
comprehensively compares our sensor with previously reported paper-based pressure sensors in the aspect
of sensitivity and detection range [22-32] , showing our sensor yields the best sensing performance. Due to the
readily availability and simplicity of fabrication, our sensor can meet the industrial and commercial mass
production requirements. Figure 3E measures the response and recovery times to be ~25 and 45~ ms,
respectively, by taking the average value of five consecutive waveforms [Supplementary Figure 6], which
indicates the quick detection capability for real-time monitoring. Figure 3F shows good dynamic behavior
under escalating pressures and different frequencies with distinct and consistent response peaking curves,
ensuring its ability to detect different physiological signals and body motions in a fast and accurate way.
Figure 3G shows the cyclic pressure detection under different compression rates from slow (~1 mm·min )
-1
to fast (~10 mm·min ) by the same 5 kPa, which demonstrates a stable dynamic sensing behavior. The long-
-1
term stability is also verified by repeatedly loading and unloading a pressure for 10,000 cycles, and stable
response curves with no drafting are observed [Supplementary Figure 7].

Application of sensor unit on airbag pillow for snoring monitoring
Due to the good breathability and excellent sensing performance, the flexible sensor can be worn on
different parts of the human body to detect various human activities such as finger bending, hand griping,
throat swallowing, mouth breathing, and wrist pulse beating [Supplementary Figure 7]. Especially the can
sensor can be used to detect joint bending for posture and gesture monitoring. During practical wearable
usage, because the sensor is assembled by multiple layers, severe tearing with excessive sheering force
should be prohibited to avoid delamination. In addition to being attached to the human body, the
developed flexible and breathable sensor can also be integrated into an airbag pillow for sleep monitoring
[Figure 4A], with detection of the airbag internal pressure using a home-built signal acquisition and
processing system [Figure 4B]. The relative current of the sensor and the internal pressure of the airbag can
be simultaneously acquired [Figure 4C], which is used as a reference for subsequent experiments.

Figure 4D shows that the activities of lying down and sitting up can be detected through the jumping and
dropping of the response waveform due to contacting of the head with the sensor. Besides, when lying
down, the activity of snoring can also be detected because it applies more pressure on the sensor during
snoring. Furthermore, the amplitude and duration of the snoring waveform can be used to distinguish
different snoring modes. The explanation can be described as follows: after lying down on the bed, the back