RESEARCH ARTICLE

www.advancedscience.com

Temperature-Triggered Adhesive Bioelectric Electrodes with

Long-Term Dynamic Stability and Reusability
Huiting Lai, Yan Liu, Yin Cheng,* Liangjing Shi, Ranran Wang,* and Jing Sun

Bioelectric electrodes with low modulus and high adhesion have been
1. Introduction
intensively pursued, as they afford conformal and strong bonding at With the development of wearable de-
skin-electrode interface to improve the fidelity and stability of vices and wireless communication, real-
electrophysiological signals. However, during detachment, tough adhesion time monitoring of bioelectric signals, such
as electrocardiogram (ECG), electromyog-
can cause pain or skin allergy; worse still, the soft electrodes can suffer
raphy (EMG), and electroencephalogram
damage due to excessive stretch/torsion, hampering long-term, dynamic, and (EEG) has been realized to help better un-
multiple uses. Herein, a bioelectric electrode is proposed by transferring silver derstand the pathological and physiologi-
nanowires (AgNWs) network to the surface of bistable adhesive polymer cal conditions of the human body.[1–4] In
(BAP). The phase transition temperature of BAP is tuned to be slightly below addition, recognition of EMG or EEG sig-
nals can assist disabled people to type[5]
skin temperature at 30 °C. Triggered by skin heat, the BAP electrode achieves
or grasp objects,[6] and also provide a wide
low modulus and high adhesion within seconds, allowing robust range of applications in human-computer
skin-electrode interface under dry, wet, and body-moving conditions. Ice bag interaction.[7,8] These applications require
treatment can dramatically stiffen the electrode and reduce the adhesion, bioelectric electrodes to be able to accu-
which allows painless detachment and avoids electrode damage. Meanwhile, rately identify complex and weak bioelectric
the AgNWs network with biaxial wrinkled microstructure remarkably signals, sustain reusable and long-duration
monitoring, and remain stable against dy-
promotes the electro-mechanical stability of the BAP electrode. The BAP
namic activities and sweat disturbances.
electrode successfully combines long-term (7 days) and dynamic (body Bioelectric electrodes can generally
movements, sweat, underwater) stability, reusability (at least ten times), and be divided into wet, semi-dry, and dry
minimized skin irritation during electrophysiological monitoring. The high electrodes.[9] Gel electrodes are typical wet
signal-to-noise ratio and dynamic stability are demonstrated in the application electrodes. Commercial Ag/AgCl gel elec-
trodes can be used to collect high-quality
of piano-playing training.
bioelectric signals under static and short-
time events. However, the easy drying of the
electrolyte gel layer reduces its electrical conductivity and inter-
facial adhesion with the skin during long-term use, resulting
H. Lai, Y. Liu, Y. Cheng, L. Shi, R. Wang, J. Sun in signal distortion.[10] The water retention of the gel electrode
State Key Laboratory of High Performance Ceramics and Superfine
Microstructure can be improved by introducing hygroscopic materials such as
Shanghai Institute of Ceramics glycerol,[11] phytic acid,[12] and LiCl salts[13] to extend its stable
Chinese Academy of Sciences operation life up to 72 h at most. However, gel electrodes that
1295 Ding Xi Road, Shanghai 200050, China can work reliably for more than 3 days are rarely reported. In ad-
E-mail: chengyin@mail.sic.ac.cn; wangranran@mail.sic.ac.cn
dition, due to the presence of hydrophilic groups in these gels,
H. Lai, Y. Liu
Center of Materials Science and Optoelectronics Engineering
they tend to swell and lose the robust bonding with skin when
University of Chinese Academy of Sciences exposed to water, which limits their practical use for underwa-
19 Yuquan Road, Beijing 100049, China ter and sweating events.[14,15] At the same time, issues such as
R. Wang allergic reactions caused by long-term adhesion also need to be
School of Chemistry and Materials Science addressed.[16]
Hangzhou Institute for Advanced Study Semi-dry electrodes can store and release electrolyte solution
University of Chinese Academy of Sciences
1 Sub-lane Xiangshan, Hangzhou 310024, China (such as NaCl solution) from reservoirs, which solve the prob-
lem of water loss in long-term monitoring.[17–19] The semi-dry
The ORCID identification number(s) for the author(s) of this article EEG electrodes with sponge structure as reservoirs have low skin-
can be found under https://doi.org/10.1002/advs.202300793 electrode impedance and high accuracy of EEG signals, which is
© 2023 The Authors. Advanced Science published by Wiley-VCH GmbH. comparable to commercial gel electrodes.[20] This stems from the
This is an open access article under the terms of the Creative Commons high flexibility of the sponge structure that allows the electrode to
Attribution License, which permits use, distribution and reproduction in
any medium, provided the original work is properly cited.
bypass the hair to form a conformal contact with the scalp. How-
ever, the reservoir of semi-dry electrodes is usually bulky and the
DOI: 10.1002/advs.202300793

Adv. Sci. 2023, 10, 2300793 2300793 (1 of 12) © 2023 The Authors. Advanced Science published by Wiley-VCH GmbH
www.advancedsciencenews.com www.advancedscience.com

electrolyte needs to be manually loaded, which limits the conve- trodes, the BAP electrodes provide bioelectric signals of higher
nient daily use. quality, especially under harsh interferences such as body move-
Gel-free dry electrodes can overcome the problem of signal ments and sweating. They impressively maintain stable signal
distortion caused by gel dehydration. However, the mechanical throughout 7 days of continuous ECG monitoring, or during 10
mismatch between the soft skin and relatively rigid dry electrode cycles of electrode reusing. By virtue of the combined high sig-
usually results in high contact impedance. Worse still, their sig- nal/noise ratio and dynamic stability in surface EMG monitor-
nals are highly susceptible to motion artifacts. An effective strat- ing, the BAP electrodes also show advantages in finger motion
egy to address these issues is to promote the flexibility of the recognition, and succeed in piano playing monitoring, including
dry electrodes. For example, fabric electrodes can be highly flex- the recognition of five-finger playing and music score, and the
ible, which are conducive to long-term monitoring.[21,22] Unfor- identification of articulation and dynamics.
tunately, their contact impedance is higher than that of commer-
cial gel electrodes due to poor conformity with skin and the small
contact area of porous fabric structures. The flexibility of dry elec- 2. Results and Discussion
trodes can also be improved by reducing the thickness or Young’s
modulus.[23] Zhao et al.[24] prepared an ultra-thin dry electrode 2.1. The Design of the BAP Electrode
with a thickness of about 100 nm, and the enhanced conforma-
bility to skin contributed to higher signal-to-noise ratio (SNR) The BAP electrode consists of an AgNWs network with a wrin-
of 23 ± 0.7 dB in comparison with 19 ± 0.5 dB of the Ag/AgCl kled microstructure and a temperature-sensitive BAP substrate
electrode. Tang et al.[25] introduced glycerol and polysorbate into (about 0.5 mm) (Figure 1A). The preparation process is shown
PEDOT: PSS-based electrode to reduce the Young’s modulus to in Figure S1, Supporting Information. The BAP is formed by
80 kPa (similar to that of skin), giving rise to bioelectric signal UV polymerization of octadecyl acrylate (SA), tetradecyl acrylate
with high SNR (≈35 dB). (TA), and urethane diacrylate (UDA) oligomer. The temperature
Although the promoted flexibility leads to high quality signal response of BAP is attributed to the crystalline-amorphous phase
through optimized interface conformability and contact area,[26] transformation of the side chains (SA and TA). UDA is a long-
the relative movement at the skin-electrode interface still induces chain oligomer that can form a bottlebrush polymer with the side
interference signals, representing a most intractable problem es- chains. It is conducive to the ordered crystallization of the side
pecially during long-term and dynamic monitoring. Therefore, it chains, reduces the phase transition temperature range of the
is critical to maintain a robust interface bonding for high-fidelity polymer, and improves the degree of crystallization of the side
bioelectric signals.[27] One way is to chemically modify the elec- chains, making the change of modulus before and after the phase
trode material to form attractive interaction at the interface, such transition more significant.[35,38] Moreover, UDA homopolymer
as hydrogen bond, dynamic covalent bond, electrostatic interac- has a modulus of 0.827 MPa and an elongation at break of 1100%,
tion, etc.[28,29] Another approach is to raise the adhesive force which guarantees the low modulus and high ductility of BAP
by means of biomimetic micropillars or sucker structures.[30-32] films.[39] According to Figure 1B, when T > Tm (melting temper-
However, such boosted extra-strong adhesion inflicts irritation, ature), the polymer matrix becomes soft as the steric hindrance
pain, or even allergy to the skin during detachment.[33,34] Besides, effect of side chains is diminished. As a result, the BAP exhib-
electrodes with low modulus and high adhesion are particularly ited high fluidity and dissipation characteristics, thus achieving
prone to excessive deformation when detached, which leads to low modulus and high adhesion. However, when T < Tm , SA and
irreversible damage to the electrode, making it impossible for TA transform into crystalline state, which stiffens the polymer
continuous high-quality signal collection and electrode reusabil- chains, and the fluidity and dissipation characteristics are signif-
ity. To the best of our knowledge, bioelectric electrodes that suc- icantly reduced, resulting in the increase of modulus and the de-
cessfully combine long-term (>3 days) and dynamic (body move- crease of adhesion. Ideally, Tm should be slightly less than the
ments, sweat, underwater, et al.) stability, reliable reusability, and skin temperature, so that BAP can not only quickly change into
minimized skin irritation have rarely been reported. amorphous state merely through body heat triggering for confor-
Here, we proposed a dry electrode with on-demand interfacial mal attachment of electrode on skin, but also transform instantly
adhesion switch based on bistable adhesive polymer (BAP)[35] into crystalline state through cold water or ice bag treatment to
and highly conductive silver nanowires (AgNWs) network. The achieve painless detachment and avoidance of electrode damage.
BAP electrodes can rapidly switch between two states by chang- The combination of low modulus and high adhesion at skin
ing the temperature. When triggered by skin heat, the BAP temperature contributes to a conformal and stable skin-electrode
electrodes achieve high flexibility and strong adhesion, and ad- interface, which facilitates the maintenance of low interfacial
here conformally and firmly to the skin even in wet conditions. contact impedance. The wrinkled microstructure of AgNWs per-
Through cold water or ice bag treatment, the electrodes can colation network endows the BAP electrode with stable conduc-
be easily detached from the skin due to significantly increased tivity against mechanical deformation. Therefore, the BAP elec-
modulus and reduced adhesion. Compared with other triggering trode enables the bioelectric signal collection both in long-term
methods (pH,[36] humidity,[33] magnetism,[37] etc.), temperature (Figure 1C) and dynamic (Figure 1D) conditions. The high mod-
control is more convenient and easier. A wrinkled microstructure ulus and low adhesion of BAP electrodes during detachment, as
is engineered at the BAP electrode surface to extend the stretch- well as the excellent stretchability of the wrinkled AgNWs net-
ability of the AgNWs layer for resisting the irreversible damage work, ensure the consistency of BAP electrode performance dur-
of the conductive network caused by skin deformation and elec- ing multiple reuses (Figure 1E), which assists in reducing costs
trode detachment. Compared to commercial Ag/AgCl gel elec- and electronic waste.

Adv. Sci. 2023, 10, 2300793 2300793 (2 of 12) © 2023 The Authors. Advanced Science published by Wiley-VCH GmbH
www.advancedsciencenews.com www.advancedscience.com

Figure 1. Schematic diagram of the working mechanism of BAP electrode and the advantages in bioelectric sensing. A) The schematic diagram of the
BAP electrode structure. B) Schematic diagram of the on-demand adhesion switch of BAP. The advantages of the BAP electrode in bioelectric signal
sensing: C) long-term monitoring for 7 days, D) dynamic monitoring, and E) reusability for ten times.

2.2. Mechanical and Adhesive Performance of BAP Electrodes The phase transition of BAP triggered by temperature also
causes the change in adhesive properties. The high storage mod-
As SA and TA have different melting points due to their differ- ulus at room temperature leads to significantly lowered adhesion
ent alkyl chain lengths, the copolymer BAP’s phase transition strength between BAP films (with no AgNWs coating) and skin.
temperature can be tuned to be slightly lower than skin tempera- On the contrary, the side chains in polymer networks change
ture by adjusting the ratio of SA and TA in the polymer network. from crystalline to amorphous states above the phase transition
Figure 2A exhibits the differential scanning calorimetry (DSC) temperature, which enhances the fluidity of the molecular chains
curves of BAP with different ratios of SA, TA, and UDA. When and remarkably reduces the storage modulus, endowing BAP
the mass ratio of SA: TA: UDA is 2:2:1, Tm is about 30 °C, at which films with high energy dissipation characteristics, and improv-
point the crystalline-amorphous phase transformation can be ac- ing the adhesive strength.[40] The high fluidity also allows more
tivated simply by skin heat. Therefore, we chose the BAP with adhesive functional groups such as amino-ester bonds to be ex-
the ratio of 2:2:1 for the preparation of bioelectric electrode in posed to form sufficient interactions with the skin, further im-
the following study. The phase transition is further confirmed proving adhesion. In addition, the low-modulus ensured confor-
by the large variation of the storage modulus as a function of mal adhesion to skin without visible gaps, resulting in increased
temperature (Figure 2B). The storage modulus of BAP decreased contact area (Figure 2D). The adhesive strength of BAP films
from 1682–2816 kPa in the crystalline state to 3.45–12.3 kPa in to pig skin was as low as 0.90 ± 0.43 kPa at 20 °C (room tem-
the amorphous state for the investigated recipes of BAP. At this perature), then soared to 71.84 ± 4.40 kPa after being heated
point, the polymer stiffness drops significantly within 30 s due to to 32 °C (skin temperature). At skin temperature, the adhesive
the narrow phase transition temperature window located slightly strength of BAP films in dry state was 71.84 ± 4.40 kPa, while it
lower than human body temperature. Furthermore, the higher decreased slightly in underwater and sweaty states, which were
fluidity of the BAP at the skin temperature gives it a higher elon- 60.83 ± 3.31 kPa and 52.01 ± 3.15 kPa. The adhesive strength of
gation. As shown in Figure 2C, the elongation of BAP at break in BAP films still maintains a high level (>50 kPa) even in sweaty
the amorphous state (32 °C) is 697%, which is much larger than state, enabling BAP films to adhere firmly to the skin (Figure 2E).
that in the crystalline state (20 °C, 116%). This can be ascribed to the high hydrophobicity of the BAP,

Adv. Sci. 2023, 10, 2300793 2300793 (3 of 12) © 2023 The Authors. Advanced Science published by Wiley-VCH GmbH
www.advancedsciencenews.com www.advancedscience.com

Figure 2. Mechanical and Adhesive Performance of BAP electrodes. A) DSC diagram of BAP with different composition ratios of SA: TA: UDA. B)
Relationship between storage modulus and temperature of BAP with different composition ratios. C) Stress–strain curves of the BAP film at 20 and 32
°C. D) Cross-sectional SEM image of the BAP electrode-skin model at skin temperature; scale bar: 20 μm. E) Tensile adhesive strength of the BAP film
on pig skin at skin temperature and room temperature in dry, underwater, and sweaty states. F) Comparison of tensile adhesive strength between the
BAP film and commercial tapes on pig skin. G) Optical photographs of attachment (left) and detachment (right) of BAP with different Tm . H) Optical
photographs of skin when the BAP electrode was i) attached and ii) detached, with no skin redness or irritation observed. I) Tensile adhesive strength
of the BAP electrode during 10 cycles of attachment.

which hinders the infiltration of water into the interface between ing the phase transition temperature of BAP to be slightly lower
the BAP film and the target surface. Figure 2F discloses that than the skin temperature enables a convenient switch of adhe-
our BAP substrate exhibits higher adhesive strength than sev- sive strength. For Tm higher than skin temperature, the body
eral kinds of commercial adhesive tapes, which qualifies the BAP heat fails to trigger the high adhesion of BAP. For Tm much
films as reliable adhesive tapes in practical applications. Adjust- lower than skin temperature, the ice bag treatment is not able to

Adv. Sci. 2023, 10, 2300793 2300793 (4 of 12) © 2023 The Authors. Advanced Science published by Wiley-VCH GmbH
www.advancedsciencenews.com www.advancedscience.com

provide a quick adhesion degradation to allow painless detach- straining has low resistance (8.9 Ω), which ensures the high con-
ment and avoid electrode damage (Figure 2G). Importantly, af- ductivity of the electrode, while the electrode obtained by uniaxial
ter the AgNWs network transfer to BAP film, the obtained BAP pre-straining has high resistance (1042.6 Ω) (Figure S9, Support-
electrode inherited high adhesive strength with no obvious de- ing Information). Moreover, the pre-straining degree of BAP elec-
crease (Figure S2, Supporting Information). Due to the semi- trodes also has a significant effect on the ability of the conductive
embedded structure of AgNWs at the polymer surface and the layer to resist deformation. Figure 3B shows the tensile strain
high fluidity of the substrate, the surface of the BAP electrode dependence of relative resistance change of BAP electrodes with
is still polymer-dominated, which is conducive to the mainte- different pre-straining degrees. The resistance of the 0%, 25%,
nance of adhesion strength. High adhesion ensures no relative and 50% pre-straining electrodes started to dramatically increase
displacement at the electrode-skin interface during skin defor- (ΔR/R0 > 5) under 25%, 35%, and 70% tensile strain, while the
mation (Figure S3, Supporting Information), which is of great conductive network was considerably damaged (ΔR/R0 > 50) un-
significance for dynamic signal acquisition. After peeling off the der 30%, 60%, and 90% tensile strain. At 50% tensile strain, the
electrodes from the skin, the texture of the skin can be com- ΔR/R0 of the BAP electrode was only 0.6 when the pre-strain
pletely reproduced, indicating the intimate topographic match- was 50%, while the ΔR/R0 of the BAP electrode was 23.61 when
ing at the skin surface (Figure S4, Supporting Information). Al- the pre-strain was 25%, and the electrode with pre-strain of 0%
though the pressure-sensitive adhesive layer of the commercial lost its conductivity. Higher pre-strain results in higher wrinkle
Ag/AgCl electrode also has high adhesive strength to skin, it density, which can store larger strain along the pre-strain direc-
will induce skin irritation and discomfort (Figure S5, Supporting tion (Figure 3C). These wrinkles gradually flatten and release
Information). The adhesion strength of the electrode decreased stress when the electrode is subjected to tensile forces, which
greatly after cold water or ice bag treatment, which facilitated easy inhibits crack generation and imparts stable conductivity dur-
removal from the skin without causing any pain, irritation, or ing the electrode deformation. Figure S10, Supporting Informa-
skin redness. (Figure 2H). What’s more, after 10 cycles of attach- tion illustrates SEM diagrams of the BAP electrode with differ-
ing and peeling-off, the adhesive strength of the BAP electrode ent pre-straining ratios at 50% tensile strain. There were numer-
did not go through appreciable degradation, still providing a high ous penetrating fractures in the 0% pre-straining electrode, while
level of around 60 kPa (Figure 2I) for potential multiple reuses. a few penetrating cracks in the 25% pre-straining electrode. In
the 50% pre-straining electrode, only the wrinkled microstruc-
ture was partly flattened, and no penetrating fractures were dis-
2.3. Electrical Performance of BAP Electrodes covered. Therefore, the BAP electrode with 50% biaxial pre-strain
was adopted in our work unless otherwise specified. Because of
Besides reliable interface adhesion, the high conductivity of the the low initial resistance of the BAP electrode (1.79 Ω), the resis-
electrode is another key factor to guarantee the acquisition of tance was still kept at a low level of 3.56 Ω (ΔR/R0 = 0.98) after
high-quality bioelectric signals. AgNWs-based percolation net- 2000 cycles of stretching, which can meet the needs of bioelec-
work was used as the conductive layer in this work consider- tric monitoring (Figure 3D). After cyclic stretching, the wrinkled
ing the high conductivity, good flexibility, and stretchability.[41,42] structure is well maintained and there is no observable network
Figure 3A presents the relationship between the sheet resistance disconnection in the AgNWs network (Figure S11, Supporting
of BAP electrodes and the dosage of AgNWs. The sheet resistance Information). The resistance of the electrode with wrinkled mi-
of BAP electrodes decreased with increasing AgNWs dosage, and crostructure formed by 50% pre-straining was kept stable when
it reached as low as 0.473 Ω □−1 at 67 μg cm−2 . As the dosage con- the electrode was attached and detached from the skin for 10 cy-
tinues to increase, the electrode conductivity leveled off, and the cles, while the electrode without pre-straining underwent a sharp
excessive amounts of AgNWs also increased the risk of exfolia- rise of resistance merely after 3 cycles (Figure 3E). Therefore,
tion. Therefore, the optimized amount of AgNWs in this work the wrinkled microstructure also favors the reusability of BAP
was set at 67 μg cm−2 . The DSC diagram of the BAP film and electrodes. SEM image proves no obvious damage to the conduc-
the BAP electrode shows that AgNWs covering the surface of the tive network of the BAP electrode after 10 cycles of repeated use
BAP substrate do not significantly affect the behavior of the BAP (Figure S12, Supporting Information). In addition, the high flu-
phase transition (Figure S6, Supporting Information). idity of the BAP substrate activated at high temperature leads to
In order to endow the AgNW-based conductive network with the configuration of AgNWs network as being semi-embedded
high tolerance to stretch caused by skin deformation (generally in the polymer substrate, which anchors the conductive layer for
less than 50%) in daily activities, a biaxial pre-straining strategy enhanced stability (Figure S13, Supporting Information). In tape
was adopted to construct wrinkled microstructure in the conduc- test, the ΔR/R0 of the BAP electrode was only 0.5 after the elec-
tive layer. Figure S7, Supporting Information shows the optical trode was pasted with tape for 50 times (Figure 3F). After 20 days
microscopy photographs of the BAP electrode at 20 and 32 °C, storage in air, resistance increase of the BAP electrode was negli-
which confirms that the wrinkled microstructure is well main- gible, testifying its excellent environmental stability (Figure 3G).
tained in the amorphous and crystalline states. Uniaxial pre- The dynamic and static stability of the conductance provides an
straining leads to a large number of cracks perpendicular to the important basis for long-term stable bioelectric monitoring of
pre-straining direction in the conductive network due to Pois- BAP electrodes.
son effect (Figure S8, Supporting Information). By contrast, the Low skin-electrode contact impedance is another guarantee
wrinkled electrode of biaxial stretching maintains the integrity for the high quality of bioelectric signals. Figure 3H indicated
of the conductive network of AgNWs and better resists multi- that the skin-electrode contact impedance of the BAP electrodes
directional skin strain. The electrode obtained by biaxial pre- was lower than that of the Ag/AgCl electrodes at all testing fre-

Adv. Sci. 2023, 10, 2300793 2300793 (5 of 12) © 2023 The Authors. Advanced Science published by Wiley-VCH GmbH
www.advancedsciencenews.com www.advancedscience.com

Figure 3. Electrical Performance of BAP electrodes. A) Relationship between AgNWs dosage and electrode sheet resistance. B) The relative resistance
variation versus stretch strain of BAP electrodes with different pre-strains (0%, 25%, and 50%). C) SEM images of BAP electrodes surface with different
pre-strains i) 0%, ii) 25%, iii) 50%, scale bar: 200 μm. D) The relative resistance variation of the BAP electrode during 2000 cycles of stretching at strain
of 40%. E) The relative resistance variation of BAP electrodes (pre-strains of 0% and 50%) during 10 consecutive attach/detach cycles. F) The relative
resistance variation of BAP electrode during 50 cycles of tape test. G) The relative resistance variation of BAP electrode during 20 days storage in air. H)
On-skin interface impedance of the BAP electrode and Ag/AgCl electrode under dry and sweaty conditions. I) Skin-electrode interface impedance of the
BAP electrode and Ag/AgCl electrode under dry and sweaty conditions at 1 Hz.

Adv. Sci. 2023, 10, 2300793 2300793 (6 of 12) © 2023 The Authors. Advanced Science published by Wiley-VCH GmbH
www.advancedsciencenews.com www.advancedscience.com

quencies in both dry and sweat states. Especially in the low fre- addition to underwater environments, BAP electrodes can offer
quency range wherein the bioelectric signal is located, the con- significant advantages in body-moving situations such as walking
tact impedance of the BAP electrodes has a significant advantage. or jogging. During walking, the P-QRS-T peaks of ECG signals
For example, as shown in Figure 3I, the skin-electrode contact acquired with the BAP electrode were distinguishable, whereas
impedance of the commercial Ag/AgCl electrodes was 2410.61 ± the T peaks of the ECG signal measured with the Ag/AgCl elec-
842.54 kΩ cm2 at 1 Hz in dry state, while that of the BAP elec- trode failed to be distinguished due to interference induced by
trodes was 304.71 ± 18.02 kΩ cm2 , which was only about one- motion artifact (Figure 4D). During jogging, the acquisition of
eighth of that of the commercial electrodes. In a sweaty state, the ECG signals was violently disturbed by the shaking of the test
contact impedance of both kinds of electrodes decreased. Even box and conducting wires, so that P and T peaks were completely
though, the contact impedance of the BAP electrodes (147.70 ± submerged in strong noise. Therefore, the raw signals were band-
8.69 kΩ cm2 @ 1 Hz) was still much smaller than that of com- pass filtered at 10 Hz – 100 Hz. After filtering, Figure 4E shows
mercial Ag/AgCl electrodes (673.93 ± 367.42 kΩ cm2 @ 1 Hz). the P and T peaks of the ECG signals measured by the BAP and
In addition, when the electrodes were attached to the sweaty commercial Ag/AgCl electrodes can be recognized clearly. Body
and moving arm, the standard deviation of contact impedance movement is usually accompanied by sweating, another main
of the BAP electrodes at 1 Hz was only 8.69 kΩ cm2 , which factor that causes signal interference. Here, simulated sweat (nor-
was much smaller than that of the commercial Ag/AgCl elec- mal saline) was used to evaluate the anti-sweat interference abil-
trodes (367.42 kΩ cm2 ), demonstrating the superb stability of ity of the two electrodes. When soaked in simulated sweat, the
the BAP electrodes under dynamic conditions. In general, we at- ECG signals of the Ag/AgCl electrodes underwent severe distor-
tribute the low and stable skin-electrode contact impedance un- tion, and the P and T peaks were difficult to be differentiated even
der both static and dynamic (moving or sweating) conditions to after band-pass filtering. In sharp contrast, the signals acquired
the remarkable combination of high conductivity, low modulus, by the BAP electrodes were almost unaffected and still afforded a
and high adhesion of the BAP electrode triggered by skin tem- clear P-QRS-T waveform (Figure 4F). The comprehensive perfor-
perature. Consequently, they demonstrate great application po- mance comparison under various conditions verified the advan-
tential in the acquisition of bioelectric signals in various daily tage of our BAP electrodes over commercial gel electrodes for
activities. daily ECG monitoring, especially when it comes to challenging
outdoor activities.
Remarkably, the long-term electrical and adhesive stability of
2.4. Bioelectric Signals Monitoring Using BAP Electrodes the BAP electrodes collectively enable long-term and continuous
signal monitoring. During the 7-day ECG monitoring test, the
2.4.1. Monitoring of ECG Signals BAP electrodes did not experience detachment or damage during
daily activities, and the ECG signals remained stable throughout,
Heart disease is characterized by abruptness, intermittency, and without baseline drift or peak loss (Figure 5A). Figure 5B presents
permanence, and patients require long-term continuous treat- the sensitivity of the ECG signal on each day, which maintains a
ment and monitoring. Hence, long-term ECG monitoring is of high level of around 0.25 and a low relative standard deviation
great clinical significance. It requires electrodes to maintain both of only 11.5%. Simultaneously, the non-solvent dry nature of the
high accuracy and robust stability during signal acquisition in BAP electrode exempts itself from the risk of water loss, making
various daily situations. Figure 4A compares the ECG signals it superior to gelled electrodes in long-term use.
recorded by BAP electrodes and Ag/AgCl gel electrodes in static Long-term monitoring of bioelectrical signals requires good
state. Both waveforms showed clear P-QRS-T peaks. The sensi- biocompatibility of electrodes. Figure S14, Supporting Informa-
tivity of ECG electrodes can evaluate the quality of ECG signals, tion shows the live/dead fluorescence staining images of NIH3T3
which is defined as the relative voltage ratio of the measured T cells after 24 h of culturing. There is no significant difference be-
peak to R peak (T/R value). A higher T/R value indicates a higher tween the cells cultured with and without BAP electrodes, indi-
accuracy of ECG signals. The sensitivity of the BAP electrodes cating that BAP electrodes are not cytotoxic. Moreover, the BAP
was 0.25, higher than that of the Ag/AgCl electrodes (0.22) dur- electrode was attached to the skin for 7 days, and there was no
ing static ECG acquisition (Figure 4B). Therefore, the BAP elec- skin redness or allergy after removing it (Figure S15, Supporting
trode promises more accurate signals for clinical heart disease Information). Therefore, the BAP electrode has good biocompat-
diagnosis. ibility and anti-allergenic properties in long-term monitoring.
In daily ECG monitoring, various activities and complex con- Moreover, after ice bag treatment, the crystalline-state BAP
ditions challenge the dynamic stability of the signal acquisition. electrode exhibited not only a considerably decreased skin ad-
Figure 4C compares the underwater ECG signals of BAP elec- hesive strength, but also a relatively stiffer conductive layer.
trodes and Ag/AgCl electrodes. For underwater arm motion and Both of these effects effectively suppress the otherwise excessive
water flow impact, the BAP electrode could acquire stable ECG stretch/distortion on BAP electrodes which could damage the
signals with clear P-QRS-T peaks, which verified their applica- conductive network. As a result, BAP electrodes allow multiple
tion in bioelectric monitoring during dynamic underwater activ- reuses, which contribute to reducing costs and electronic waste.
ities such as swimming and bathing. By contrast, for Ag/AgCl As shown in Figure 5C, the ECG signals recorded by BAP elec-
electrodes, the P and T peaks of ECG signals attenuated or even trodes were stable after repeated use on the skin for ten times.
disappeared when swinging the arm underwater. The impact of High retention of mechanical properties and electrical conductiv-
water flow even brought about the delamination of the Ag/AgCl ity during cyclic adhesion and debonding is responsible for such
electrode from the skin and the interruption of ECG signals. In robust reusability.

Adv. Sci. 2023, 10, 2300793 2300793 (7 of 12) © 2023 The Authors. Advanced Science published by Wiley-VCH GmbH
www.advancedsciencenews.com www.advancedscience.com

Figure 4. BAP electrodes for ECG signal monitoring in static and dynamic states. A) ECG signals and B) T/R values recorded by BAP and Ag/AgCl
electrodes in static state. C) ECG signals recorded by BAP electrodes and Ag/AgCl electrodes during arm motion underwater (left) and water flow
impact (right) underwater conditions. D) ECG signals recorded by BAP electrodes and Ag/AgCl electrodes during walking. ECG signals recorded by BAP
electrodes and Ag/AgCl electrodes during E) jogging and F) jogging with simulated sweat.

Adv. Sci. 2023, 10, 2300793 2300793 (8 of 12) © 2023 The Authors. Advanced Science published by Wiley-VCH GmbH
www.advancedsciencenews.com www.advancedscience.com

Figure 5. BAP electrodes for long-term and reusable ECG signal monitoring. A) ECG signals and B) T/R values monitored continuously by BAP electrodes
for 7 days. C) ECG signals monitored by BAP electrodes during 10 consecutive attach/detach cycles.

2.4.2. Monitoring of surface EMG Signals electrodes were placed in the position shown in Figure 6C-i for
fingering recognition. When playing the piano, the thumb, index
Surface EMG (sEMG) records the electrical activity of muscles finger, middle finger, ring finger, and little finger exclusively
from the surface of the skin, which is non-invasive and easy to correspond to five notes, do (1), re (2), mi (3), fa (4) and so (5)
operate, and has broad prospects in the field of human-machine respectively. The original signals of sEMG are complex and diffi-
interaction.[43,44] Precise motion recognition requires the signal cult to compare directly. Therefore, we need to analyze the sEMG
accuracy to be as high as possible. The sEMG signals of BAP signals in the time domain and the frequency domain to more
electrodes and Ag/AgCl gel electrodes were recorded during fist intuitively and accurately show the signal differences of different
clenching (Figure 6A). The SNR of the sEMG signals recorded by fingers. We selected RMS as the time domain eigenvalue, which
the BAP electrodes was 18.773 dB, which was higher than that represents the amplitude of the sEMG signal and reflects the
recorded by the Ag/AgCl gel electrodes (17.932 dB). The base- degree of muscle activation. The median frequency (MF) is an
line potential represents the noise level, which can characterize eigenvalue in the frequency domain, which represents the speed
the signal quality from another aspect. As shown in Figure 6B, of muscle contraction. Compared with other frequency-domain
the root mean square (RMS) of the baseline potential recorded eigenvalues, MF has stronger anti-noise performance.[45] Ac-
by the BAP electrodes was only 3.35 μV, which was lower than cording to Figure 6C-ii, we extracted the RMS and MF values of
that recorded by the Ag/AgCl gel electrodes (5.02 μV). Both high sEMG signals when five fingers played, so as to realize the recog-
SNR and low baseline potential represent better quality of sEMG nition of different playing fingers. Thus, we can reconstruct the
signals recorded by BAP electrodes. Therefore, the BAP elec- music score by identifying the playing fingers from the sEMG
trodes are more promising to implement sEMG-based human signals (Figure 6C-iii). Further, for different piano articulations
motion recognition with high precision. Piano playing is selected (the specific way to strike a piano key to make a desired sound),
here as a representative application for it demands delicate finger there are significant differences in the duration of sEMG signals
movements. (Δt), as shown in Figure 6D-i. Under the same playing rhythm,
For piano-learning beginners, finger movement recognition the Δt produced by staccato is the shortest, followed by non-
can assist in efficient practice. Although strain sensors attached legato, and legato is the longest, as shown in Figure 6D-ii. In
to the finger can give high recognition accuracy, they inevitably this way, the sEMG signals measured by BAP electrodes can dis-
limit finger movements and affect proprioception during prac- tinguish different types of piano articulation, including staccato,
tice. The BAP electrodes attached to the arm to collect sEMG non-legato, and legato. Also, Figure 6D-iii indicates that the RMS
signals can avoid such interference to the piano playing. The BAP values of the sEMG signals of the five fingers are significantly

Adv. Sci. 2023, 10, 2300793 2300793 (9 of 12) © 2023 The Authors. Advanced Science published by Wiley-VCH GmbH
www.advancedsciencenews.com www.advancedscience.com

Figure 6. BAP electrodes for sEMG signal monitoring. A) The sEMG signals and SNR value of BAP electrodes and Ag/AgCl electrodes in fist clenching.
B) The sEMG signal baseline and RMS value of BAP electrodes and Ag/AgCl electrodes. C) i) Schematic diagram of BAP electrodes in piano fingering
identification. ii) RMS and MF of the sEMG signals of each finger during playing. iii) The sEMG signals and the recognized music score monitored by
BAP electrodes. D) i) The sEMG signals collected under different types of piano articulation and ii) Δt and iii) RMS of sEMG when playing piano with
each finger. E) i) The sEMG signals at different music dynamics, ii) statistical distribution results of RMS, MPF, and MF extracted from the sEMG signals
corresponding to different music dynamics (pianissimo, mezzo forte, and fortissimo).

Adv. Sci. 2023, 10, 2300793 2300793 (10 of 12) © 2023 The Authors. Advanced Science published by Wiley-VCH GmbH
www.advancedsciencenews.com www.advancedscience.com

different when playing the piano under staccato, non-legato, suction filtration. The BAP film was pasted in the center of an Ecoflex film.
and legato. Consequently, the recognition of the five fingers The Ecoflex film was biaxially stretched, and the BAP film was driven to
can be realized under different piano articulations. Different produce 50% pre-straining. The AgNWs network on the PTFE membrane
was transferred to the BAP film, and the BAP electrode with wrinkled mi-
playing rhythms can also be identified through the intervals
crostructure was obtained after unloading. The process of pre-straining
of the sEMG signals (Figure S16-i, Supporting Information). and unloading was completed at 40 °C. The BAP electrode was connected
Due to the instantaneity of sEMG signal acquisition, different to the external circuit with copper conductive tape, and silver paste was
sEMG intervals represent different playing rhythms. The RMS coated on the interface between the two to ensure the circuit ran.
values of the sEMG signals generated by five-finger playing still Characterization and Electrophysiological Signal Monitoring: The
remain significantly different in different playing rhythms, thus crystalline-amorphous phase transformation temperature of polymer was
measured with a differential scanning calorimeter (DSC200F3, Netzsch,
it still has strong discriminability for five fingers (Figure S16-ii,
Germany) in the range of −25 to 70 °C at a rate of 15 °C min−1 . The
Supporting Information). In addition, the sEMG signal can temperature–storage modulus curves of polymer were characterized
also identify different music dynamics (defined as pianissimo, by a rheometer (Anton Paar, MCR301, Germany) using a temperature-
mezzo forte, and fortissimo). During the experiment, the piano scanning oscillation mode with a temperature range of 0 to 60 °C, a
player maintained the playing gesture, hitting the piano key with heating rate of 1.8 °C min−1 , a frequency of 1 Hz, and a strain of 1%.
only one finger while the other fingers remained relaxed. In The universal tensile testing machine (CMT6103, MTS Systems, China)
this case, only the muscles corresponding to the playing finger was used to test the stress–strain curves of BAP films at a tensile rate
of 0.3 mm s−1 and a standard distance of 15 mm. The tensile adhesive
produce sEMG signals, and the signals corresponding to the strength was measured by the force gauge (MARK-10, USA), and the
other fingers are ignored. Different music dynamics of the ring moving rate of the force gauge was 500 mm min−1 . The tensile adhesion
finger were chosen as examples in Figure 6E-i. When RMS, MF strength was determined by the following formula: P = -F/S, where F is
and mean power frequency (MPF) feature values were selected the peak force (negative value) read by the force gauge during movement
in the time domain features, it can be seen that the feature and S is the area of the BAP film or tape. The field emission scanning
sets corresponding to pianissimo, mezzo forte, and fortissimo electron microscope (SU8200, Hitachi, Japan) was used to characterize
the surface structure of the BAP electrode and its interface with the pig
are located in distinct zones in the triaxial coordinate system
skin model. The pigskin model was prepared by PDMS using the reverse
(Figure 6E-ii). mold method.
A low impedance surface impedance instrument (MCP-T360, Japan)
was used to measure the sheet resistance of the electrodes with different
3. Conclusion dosages of AgNWs. The electrochemical station (Autolab, Germany) was
used to analyze the electrode resistance changes with strain, the cycling
In summary, this study proposed a temperature-triggered bio-
stability of electrode resistance, and the skin-electrode contact impedance
electric electrode with on-demand adhesion and modulus based spectrum. In the cyclic stability test, the standard distance was set at 5 cm,
on BAP and AgNWs networks. The electrode achieved high ad- the stretch-recovery rate was 0.1 mm s−1 , and the test temperature was
hesion to the skin and painless detachment simultaneously. Be- 32 °C. During the skin-electrode contact impedance test, the electrodes
yond this, ten times of reuse and long-term ECG monitoring were attached to the skin of the forearm, the electrode spacing was 4 cm,
for 7 days without interference from human movements, sweat- the voltage was 0.01 V, and the frequency range was 0.1 Hz–105 Hz. The
ing, and bathing were also realized. Compared with the commer- long-term electrical stability of the electrode and the interface stability be-
tween the conductive network and the substrate were tested by a multi-
cial Ag/AgCl gel electrode, the BAP electrode has a higher SNR meter (FLUKE 18B+, China).
(18.773 dB) and a lower baseline noise (3.35 μV) in sEMG moni- The ECG signals were recorded by an ECG monitor (Heal Force PC-
toring, which enable them to facilitate smart training on piano 80B, China), and the electrodes were attached to the left wrist, right wrist,
playing. These results demonstrate the great potential of BAP and the inside of the left ankle. The sEMG signals were monitored by a
electrodes in the fields of long-duration/dynamic health moni- surface EMG collector (ZJE-II, China) at a sampling frequency of 1000 Hz.
toring and diagnostics, and high-precision human-computer in- Matlab 2018a and Origin 2020 software were used to process the ECG and
sEMG signals. The skin was wiped with alcohol before the electrodes were
teractions such as prosthetic manipulation and emotion recogni-
attached.
tion. In vitro cell experiments were performed as follows: the 10 mm diame-
ter BAP electrodes were sterilized by high temperature and pressure steam
(121 °C for 20 min) and placed in a 24-well plate, and NIH3T3 cells were
seeded on the surface of the materials. The cells were incubated for 24 h at
4. Experimental Section 37 °C in a 5% CO2 incubator. Cells in 24-well plates were washed with PBS
Materials: SA (Sigma Aldrich), TA (TCI), and DMPA (Sigma Aldrich) to remove excess serum. The staining solution (BestBio, BB-4126) was
were purchased from Greagent (Shanghai Titan Scientific Co., Ltd.). UDA added for cell staining, and the cells were incubated at room temperature
oligomer (CN9021NS, Sartomer) was obtained from Chemelite Inc. The in the dark for 15 min. Then the cells were observed and photographed
AgNWs solution (CST-NW-S70, 10 mg mL−1 , average diameter of 70 nm, using a fluorescence microscope (Zeiss, Axio Vert.A1).
length of 30 ± 5 μm, solvent was IPA) was purchased through XFNANO
Corporation. Ag/AgCl gel electrodes (CH50RB) were provided by Heal
Force Bio-meditech Holdings Limited. Supporting Information
Preparation of BAP Films: SA, TA, and UDA were mixed in the ratio of Supporting Information is available from the Wiley Online Library or from
2:2:1. 0.5 wt% DMPA was added to the mixture as a photoinitiator. The the author.
mixed solution was treated with ultrasound at 40 °C for 1 h to make it
evenly mixed. Remove bubbles from the mixed solution in a vacuum drying
oven. BAP films were obtained by pouring the mixed solution into the mold Acknowledgements
and UV curing for 3 min.
Fabrication of BAP Electrodes: The AgNWs solution was diluted with This work was financially supported by the National Natural Science Foun-
alcohol, and the AgNWs network was obtained on a PTFE membrane by dation of China (grant no 62122080, 62261136551, 52203365), the Natural

Adv. Sci. 2023, 10, 2300793 2300793 (11 of 12) © 2023 The Authors. Advanced Science published by Wiley-VCH GmbH
www.advancedsciencenews.com www.advancedscience.com

Science Foundation of Shanghai (22ZR1481700, 22dz1205000), and the [16] S. R. Yadhuraj, B. G. Sudarshan, P. S. C. Kumar, M. D. Kumar, 2nd In-
Shanghai Pujiang Program (21PJ1414800). ternational Conference on Large Area Flexible Microelectronics (ILAFM)
– Wearable Electronics, Rashtreeya Vidyalaya College of Engineering,
Bengaluru 2018.
Conflict of Interest [17] F. Wang, G. Li, J. Chen, Y. Duan, D. Zhang, J. Neural Eng. 2016, 13,
046021.
The authors declare no conflict of interest. [18] K. P. Gao, H. J. Yang, L. L. Liao, C. P. Jiang, N. Zhao, X. L. Wang, X. Y.
Li, X. Chen, B. Yang, J. Liu, IEEE Trans. Biomed. Eng. 2020, 67, 750.
[19] G. L. Li, J. T. Wu, Y. H. Xia, Q. G. He, H. G. Jin, J. Neural Eng. 2020,
Data Availability Statement 17, 051004.
[20] S. Lin, J. Liu, W. Li, D. Wang, Y. Huang, C. Jia, Z. Li, M. Murtaza, H.
The data that support the findings of this study are available from the cor- Wang, J. Song, Z. Liu, K. Huang, D. Zu, M. Lei, B. Hong, H. Wu, Nano
responding author upon reasonable request. Lett. 2019, 19, 6853.
[21] J. Dong, D. Wang, Y. Peng, C. Zhang, F. Lai, G. He, P. Ma, W. Dong,
Y. Huang, I. P. Parkin, T. Liu, Nano Energy 2022, 97, 107160.
Keywords [22] L. W. Taylor, S. M. Williams, J. S. Yan, O. S. Dewey, F. Vitale, M.
Pasquali, Nano Lett. 2021, 21, 7093.
bioelectric electrodes, dynamic monitoring, long-term monitoring, [23] Y. Chen, G. Zhou, X. Yuan, C. Li, L. Liu, H. You, Biosens. Bioelectron.
reusability, temperature-triggered adhesion switch 2022, 206, 114118.
[24] Y. Zhao, S. Zhang, T. Yu, Y. Zhang, G. Ye, H. Cui, C. He, W. Jiang, Y.
Received: February 11, 2023 Zhai, C. Lu, X. Gu, N. Liu, Nat. Commun. 2021, 12, 4880.
Revised: April 11, 2023 [25] W. Tang, Y. Zhou, S. Chen, S. Yu, Y. Yang, J. Lin, S. Yin, Y. Ma, B. Hu,
Published online: May 18, 2023 ACS Mater. Lett. 2021, 3, 1385.
[26] H. Wu, G. G. Yang, K. H. Zhu, S. Y. Liu, W. Guo, Z. Jiang, Z. Li, Adv.
Sci. 2021, 8, 2001938.
[27] Z. Jia, X. Lv, Y. Hou, K. Wang, F. Ren, D. Xu, Q. Wang, K. Fan, C. Xie,
[1] J. Li, Y. Ma, D. Huang, Z. Wang, Z. Zhang, Y. Ren, M. Hong, Y. Chen, X. Lu, Bioact. Mater. 2021, 6, 2676.
T. Li, X. Shi, L. Cao, J. Zhang, B. Jiao, J. Liu, H. Sun, Z. Li, Nano-Micro [28] C. Cui, W. Liu, Prog. Polym. Sci. 2021, 116, 101388.
Lett. 2022, 14, 132. [29] C. Xie, X. Wang, H. He, Y. Ding, X. Lu, Adv. Funct. Mater. 2020, 30,
[2] I. Mporas, V. Tsirka, E. I. Zacharaki, M. Koutroumanidis, M. 1909954.
Richardson, V. Megalooikonomou, Expert Syst. Appl. 2015, 42, 3227. [30] S. Chun, W. Son, D. W. Kim, J. Lee, H. Min, H. Jung, D. Kwon, A. H.
[3] G. Drost, D. F. Stegeman, B. G. M. van Engelen, M. J. Zwarts, J. Elec- Kim, Y.-J. Kim, S. K. Lim, C. Pang, C. Choi, ACS Appl. Mater. Interfaces
tromyogr. Kinesiology 2006, 16, 586. 2019, 11, 16951.
[4] Y. Zhang, E. R. W. van Doremaele, G. Ye, T. Stevens, J. Song, R. C. [31] H. Min, S. Jang, D. W. Kim, J. Kim, S. Baik, S. Chun, C. Pang, ACS
Chiechi, Y. van de Burgt, Adv. Mater. 2022, 34, 2200393. Appl. Mater. Interfaces 2020, 12, 14425.
[5] J. J. Norton, D. S. Lee, J. W. Lee, W. Lee, O. Kwon, P. Won, S. Y. Jung, [32] T. Kim, J. Park, J. Sohn, D. Cho, S. Jeon, ACS Nano 2016, 10, 4770.
H. Cheng, J. W. Jeong, A. Akce, S. Umunna, I. Na, Y. H. Kwon, X. Q. [33] Y. Zhang, T. H. Tao, Adv. Mater. 2019, 31, 1905767.
Wang, Z. Liu, U. Paik, Y. Huang, T. Bretl, W. H. Yeo, J. A. Rogers, Proc. [34] Y. Jiang, X. Zhang, W. Zhang, M. Wang, L. Yan, K. Wang, L. Han, X.
Natl. Acad. Sci. U. S. A. 2015, 112, 3920. Lu, ACS Nano 2022, 16, 8662.
[6] L. Pan, P. Cai, L. Mei, Y. Cheng, Y. Zeng, M. Wang, T. Wang, Y. Jiang, [35] M. Gao, H. Wu, R. Plamthottam, Z. Xie, Y. Liu, J. Hu, S. Wu, L. Wu, X.
B. Ji, D. Li, X. Chen, Adv. Mater. 2020, 32, 2003723. He, Q. Pei, Matter 2021, 4, 1962.
[7] J.-W. Jeong, W.-H. Yeo, A. Akhtar, J. J. S. Norton, Y.-J. Kwack, S. Li, S.-Y. [36] X. Zhao, Y. Liang, Y. Huang, J. He, Y. Han, B. Guo, Adv. Funct. Mater.
Jung, Y. Su, W. Lee, J. Xia, H. Cheng, Y. Huang, W.-S. Choi, T. Bretl, J. 2020, 30, 1910748.
A. Rogers, Adv. Mater. 2013, 25, 6839. [37] S. H. Wang, H. Y. Luo, C. H. Linghu, J. Z. Song, Adv. Funct. Mater.
[8] G. Ye, J. Qiu, X. Fang, T. Yu, Y. Xie, Y. Zhao, D. Yan, C. He, N. Liu, 2021, 31, 2009217.
Mater. Horiz. 2021, 8, 1047. [38] D. Zhang, E. Dashtimoghadam, F. Fahimipour, X. Hu, Q. Li, E. A.
[9] G. Li, S. Wang, Y. Y. Duan, Sens. Actuators, B 2017, 241, 1244. Bersenev, D. A. Ivanov, M. Vatankhah-Varnoosfaderani, S. S. Sheiko,
[10] J. Cao, X. Yang, J. Rao, A. Mitriashkin, X. Fan, R. Chen, H. Cheng, X. Adv. Mater. 2020, 32, 2005314.
Wang, J. Goh, H. L. Leo, J. Ouyang, ACS Appl. Mater. Interfaces 2022, [39] Z. Ren, W. Hu, C. Liu, S. Li, X. Niu, Q. Pei, Macromolecules 2016, 49,
14, 39159. 134.
[11] C. Wang, H. Wang, B. Wang, H. Miyata, Y. Wang, M. O. G. Nayeem, [40] Z. Liu, F. Yan, Adv. Sci. 2022, 9, 2200264.
J. J. Kim, S. Lee, T. Yokota, H. Onodera, T. Someya, Sci. Adv. 2022, 8, [41] L. Zhang, T. Song, L. Shi, N. Wen, Z. Wu, C. Sun, D. Jiang, Z. Guo, J.
eabo1396. Nanostruct. Chem. 2021, 11, 323.
[12] Q. Zhang, X. Liu, J. W. Zhang, L. J. Duan, G. H. Gao, J. Mater. Chem. [42] Y. Zhou, L. Qian, T. Zhang, Mater. Rev. 2020, 34, 21081.
A 2021, 9, 22615. [43] K. Li, J. Zhang, L. Wang, M. Zhang, J. Li, S. Bao, Biomed. Signal Process.
[13] Y. Niu, H. Liu, R. He, M. Luo, M. Shu, F. Xu, Small 2021, 17, 2101151. Control 2020, 62, 102074.
[14] S. Ji, C. Wan, T. Wang, Q. Li, G. Chen, J. Wang, Z. Liu, H. Yang, X. Liu, [44] C. Disselhorst-Klug, T. Schmitz-Rode, G. Rau, Clin. Biomech. 2009,
X. Chen, Adv. Mater. 2020, 32, 2001496. 24, 225.
[15] X. Su, D. Z. Hao, X. Q. Xu, X. L. Guo, Z. N. Li, L. Jiang, ACS Appl. [45] I. W. Penn, T. Y. Chuang, R. C. Chan, T. C. Hsu, Med. Sci. Sports Exercise
Mater. Interfaces 2020, 12, 25316. 1999, 31, 1834.

Adv. Sci. 2023, 10, 2300793 2300793 (12 of 12) © 2023 The Authors. Advanced Science published by Wiley-VCH GmbH