Page 102 - Read Online

Tang et al. Soft Sci. 2025, 5, 11 https://dx.doi.org/10.20517/ss.2024.62 Page 13 of 21

forward in the field of wearable health monitoring and environmental sensing. Deep learning algorithms
enhance the capability of TE sensors to not only detect but also interpret complex physiological and
environmental data with high accuracy. Zhang et al. have contributed with their deep-learning-assisted
[102]
TGH fiber sensor for self-powered in-nostril respiratory monitoring [Figure 5E] . This novel sensor
leverages the TE effect to convert temperature differences within the nasal cavity into electrical signals,
which are then processed using deep learning to identify distinct respiratory patterns with remarkable
accuracy. Using deep learning, the hydrogel fiber-based respiratory monitoring strategy could actively
identify seven respiratory patterns with an accuracy of 97.1%. Based on the sensor’s TE capabilities, the
hydrogel fiber sensor, with thermogalvanic properties, provides a non-invasive and comfortable method for
[102]
long-term monitoring of respiratory health [Figure 5F] . This is particularly significant for individuals
with conditions such as sleep apnea or asthma, where continuous monitoring is crucial for disease
management and prevention of complications.

Temperature sensor for environmental monitoring
Environmental sensing is pivotal in the realm of smart living and industrial automation, providing critical
data that informs decisions on safety, comfort, and energy efficiency [4,103,104] . The ability to accurately
monitor environmental parameters such as temperature, humidity, and gas concentrations is not just a
matter of convenience but is essential for maintaining optimal living conditions and preventing health
hazards . Advancements in material science have led to the development of innovative self-powered
[105]
temperature monitoring systems that harness TE and thermogalvanic effects, offering a new frontier in
environmental sensing technology. These systems are designed to be highly responsive and adaptive to
varying conditions, making them invaluable tools for enhancing the precision and reliability of
environmental sensing applications. Li et al. have crafted a gel electrolyte-based flexible thermogalvanic
device that incorporates the I /I redox couple. This device demonstrates remarkable temperature resilience,
-
-
3
operating effectively between -20 and 80 °C, and shows excellent resistance to drying under low vapor
pressure conditions, as illustrated in Figure 6A . This innovation allows the hydrogel to maintain its
[106]
mechanical properties even in subzero temperatures, thus enhancing its resistance to environmental
interference. The advancement is attributed to the weakening of hydrogen bonds between water molecules
through a binary solvent strategy, enhancing the device’s performance in harsh temperature environments.
The high thermal sensitivity and stretchability of the TGH make it an ideal candidate for wearable E-skins,
enabling real-time and continuous monitoring of environmental temperatures without the need for external
power sources. The hydrogel can be embedded in smart windows, known as H-windows, for self-powered
monitoring of indoor and outdoor temperatures, with the potential to trigger alarms in case of abnormal
temperature changes, such as those indicating a fire or a malfunctioning refrigeration system.

The incorporation of TE materials into wearable textiles represents an innovative approach to early fire
warning systems in harsh environments. He et al. have developed a self-powered, wearable fire warning
electronic textiles (e-textile) that leverages the TE properties of Ti C T MXene, silver nanowires (Ag NWs),
3 2 x
and aramid nanofibers (ANFs) within an aerogel fiber matrix [Figure 6B] . This e-textile, crafted through
[107]
wet spinning and weaving techniques, demonstrates a wide-range temperature sensing capability from 100-
400 °C, indicating its reliability in different temperature scenarios. The output voltage of the e-textile
increased with temperature, fitting a linear model with a high correlation coefficient (R2 = 0.957), which is
attributed to the TE properties of Mxene. This linear relationship between TE voltage and temperature
allows the e-textile to function effectively across varying temperature conditions, making it suitable for real-
time temperature monitoring during firefighting operations. Additionally, the e-textile showed only a slight
decrease in conductivity after a 500-cycle bending test, indicating its exceptional TE stability under
mechanical stress. The rapid and repeated fire warning capability of the e-textile, which can initiate an
alarm in less than 1.6 seconds upon exposure to flame, is significantly improved by the three-dimensional