Stretchable and biodegradable self-healing conductors for multifunctional electronics

Abstract

As the regenerative mechanisms of biological organisms, self-healing provides useful functions for soft electronics or associated systems. However, there have been few examples of soft electronics where all components have self-healing properties while also ensuring compatibility between components to achieve multifunctional and resilient bio-integrated electronics. Here, we introduce a stretchable, biodegradable, self-healing conductor constructed by combination of two layers: (i) synthetic self-healing elastomer and (ii) self-healing conductive composite with additives. Abundant dynamic disulfide and hydrogen bonds of the elastomer and conductive composite enable rapid and complete recovery of electrical conductivity (~1000 siemens per centimeter) and stretchability (~500%) in response to repetitive damages, and chemical interactions of interpenetrated polymer chains of these components facilitate robust adhesion strength, even under extreme mechanical stress. System-level demonstration of soft, self-healing electronics with diagnostic/therapeutic functions for the urinary bladder validates the possibility for versatile, practical uses in biomedical research areas.

A stretchable, self-healing conductor boosts the resilience and functionality of transient electronics in biomedical applications.

INTRODUCTION

Materials and devices that are designed to be flexible and stretchable have created numerous unprecedented applications in the realms of wearable health care devices (1–5), biomedical implants (6–11), tissue adhesive patches (12–14), syringe-injectable devices (15–17), human-machine interfaces (18–21), soft robotics (22–24), and virtual or augmented reality systems (25, 26). Reliable and prolonged function of these electronics substantially relies on mechanical stability of key elements such as substrates/insulators, conductors, and semiconductors against frequent external strains under various deformation modes (27, 28). In this respect, mechanically elastic conductors have been achieved through structural designs of metal electrodes in serpentine (2, 7, 29), three-dimensional configurations (30, 31), or materials engineering of intrinsically stretchable hydrogels (32, 33) and conducting polymer composites (34, 35). Although mechanical resilience can be secured through the different manners, incorporation of a self-healing chemistry can further enhance the robustness of soft electronics by ensuring the ability to recover structural and electrical integrity from unwanted mechanical damages. Several studies have developed polymers with reversible chemical bonds such as hydrogen bonds, metal-ligand interactions, and dynamic covalent bonds as effective self-healing substrates for electronic skins (36–38), bioelectronic devices (33, 39), energy devices (40, 41), and soft robotics (42). Moreover, integration of self-healing polymers with electrically conductive materials offered an opportunity to achieve self-repairable conductors. Examples include silver nanowires (AgNWs) (37, 43–45), carbon nanotubes (CNTs) (37, 46), liquid metals (36, 47) formed on/in self-healing polymers, and poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS)–based self-healing composites (48, 49). However, several issues may be raised to realize multifunctional electronic devices using these materials: (i) Physical reconstruction of conducting nanostructures (AgNWs and CNTs) led to incomplete and delayed electrical recovery; (ii) the fluidic nature of liquid metals with high surface tension resulted in difficulties associated with improper application or post-casting machining; and (iii) the poor adhesion of typical PEDOT:PSS films to polymeric substrates hindered integration with other components (50). Most of these materials exhibited strain-sensitive changes in electrical conductivity, which may not be suitable for reliable operation of soft electronics under external strains.

Biodegradable electronics—that can be disintegrated or dissolved after complete mission in a targeted period—offers versatile opportunities for various applications including implantable biomedical devices that eliminate the need for surgically removal procedures of implants and environmentally friendly consumer gadgets and monitors/sensors that minimize solid waste generation (9, 10, 51–55). While such electronics was designed to dissolve in certain time frames with biologically safe formats, reliable and stable performance over the period of service must remain the same as that of nondegradable counterparts. In this context, the combined use of self-healing capabilities may be particularly useful in dynamic and sensitive biological environments, underlining the necessity for innovative research in multifunctional electronic systems.

Here, we propose a highly stretchable and self-healing conductor for multiple components of biodegradable electronic systems. Chemical synthesis and rational materials design yield biodegradable yet self-healing elastomers and conductive composites with rapid and complete recoverable electrical properties after mechanical damages. Combination of both components results in a stretchable, biodegradable, self-healing conductor, a type of material that has never been developed before, and comprehensive evaluations validate superior properties including the ability to repair electrical conductivity/stretchability as well as strong interfacial adhesion even under ultrasonication. Demonstration of multifunctional, self-healing electronics for monitoring/treating physiological functions of the urinary bladder in vivo proves the practical applicability toward biomedical systems.

RESULTS

Stretchable, biodegradable, self-healing polymers and conductive composites

Figure 1A illustrates a soft, stretchable, biodegradable electronic tree that can monitor the humidity, temperature, and pressure distribution, with the self-healing ability to restore mechanical and electrical properties after physical damages. This device consists of two key components: (i) stretchable, self-healing conductive composite (SH-CC) and (ii) biodegradable and self-healing elastomer (fig. S1). SH-CC is composed of PEDOT:PSS, doped with biocompatible additives for enhanced mechanical and electrical properties, such as 1-butyl-1-methylpyrrolidinium bis(trifluoromethylsulfonyl)imide (P14[TFSI]), poly(ethylene glycol) (PEG), and glycerol. P14[TFSI] contributed to the stretchability of SH-CC as reported in a previous study (8), whereas PEG and glycerol promoted the reorganization of PEDOT-PEDOT conductive pathways through strong interactions between abundant hydrogen bonds and sulfonic acid groups, leading to the self-repairing feature (48). We note that PEDOT:PSS and all additives have been verified to exhibit low toxicity and biocompatibility/eco-compatibility, which supports the potential for biomedical and eco-friendly applications (8, 56–60). Synthesis using poly(lactide-co-ε-caprolactone)-diol (PLCL-diol) as a soft segment, isophorone diisocyanate (IPDI) as a hard segment, and symmetrical disulfide derivatives with OH/NH₂ end groups as chain extenders created the biodegradable, self-healing PLCL (SH-PLCL). Disulfide moieties provided room temperature (RT) self-healing property through disulfide metathesis, and urethane/urea bonds with an asymmetric alicyclic structure facilitated self-healing efficiency due to high chain mobility and hydrogen bonds while also promoting mechanical softness (61, 62). Detailed polymerization procedures, results, and analyses appear in figs. S2 to S4 and table S1. Unlike previous self-healing conductors, which typically relied on the physical reconstruction of conducting materials such as CNTs and AgNWs (37, 45), the dynamic chemical bonds in both SH-CC and SH-PLCL enable the rapid and complete recovery of mechanical/electrical properties. Furthermore, the interpenetration of polymer chains and subsequent hydrogen bonding at the interface between SH-CC and SH-PLCL ensure robust, reliable interlayer adhesion. Figure 1B exhibits the ability to completely restore the electrical performance of a tactile sensor from physical failures. The capacitive tactile sensor maintained the initial sensitivity through the self-healing process after being intentionally cut by a razor blade. Other components including temperature and humidity sensors exhibited similar behaviors (fig. S5). Such resilience was apparently revealed in real-time, continuous measurements using an array of tactile sensors with small objects (~2 g) placed on three different spots (red, green, and navy circles) (Fig. 1C and fig. S6). After electrical interconnects at the numbered regions were sequentially cut in numerical order (1 to 4), the damaged wires autonomously recovered the sensing capability within tens of seconds, highlighting the strong self-healing nature of SH-CC/SH-PLCL.

Fig. 1. Stretchable, biodegradable, and self-healing electronic tree with multifunctional sensing capabilities.

(A) Representative image of a stretchable, biodegradable, and self-healing electronic tree composed of an SH-PLCL elastomer and an SH-CC based on PEDOT:PSS with P14[TFSI] as a stretchability enhancer and PEG/glycerol as a self-healing inducer and conductivity enhancer. The inset image shows the configuration of devices implementing multimodal sensing capabilities of the electronic tree. Hydrogen bonding in SH-CC as well as disulfide exchange and hydrogen bonding in SH-PLCL facilitate the RT self-healing feature. In addition, interpenetration and subsequent hydrogen bonding between SH-CC and SH-PLCL promote strong adhesion. (B) Evaluation of the electrical performance of the tactile sensor before and after re-healing. Data are shown as mean values ± SD. n = 5 independent samples. (C) Real-time recording of capacitance changes for three small objects (~2 g). Interconnects were completely cut by a razor blade in the order of 1, 2, 3, and 4 and autonomously restored their electrical functions in ~2 min.

Biochemical, mechanical, and self-healing characteristics of SH-PLCL

The molecular structure can control the key characteristics of polymers including biochemical and mechanical properties and self-healing efficiency (63, 64). Figure 2A presents the dissolution behavior of SH-PLCLs with different molecular weights [number-average molecular weight (M_n)] of PLCL-diol (1k, 2k, and 3k) in phosphate-buffered saline (PBS; pH 7) at RT. An increase in the M_n accelerated degradation rates due to the large amount of degradable ester groups (65). The dissolution rate could be adjusted by the temperature and pH level of the solution, as previously reported (fig. S7) (8, 66). Figure 2B illustrates stress-strain curves of the SH-PLCLs. The results revealed that Young’s modulus, tensile strength, and toughness became lower as the M_n increased, while elongation at break increased (measured mechanical parameters are summarized in table S2). Such a trend primarily arose from the reduced volume ratios of hard-to-soft segments because the hard segments—composed of a rigid benzene ring and hydrogens bonds—served as physical cross-linking points (42). Additional mechanical stability studies, including hysteresis, cyclic tests, and degradation tests, for the SH-PLCL_3k elastomer are presented in figs. S8 and S9. The volume ratio also influenced the viscoelastic properties of the SH-PLCLs. Dynamic mechanical analysis (DMA) in Fig. 2C exhibited two distinct damping peaks related to the glass transition of soft (T_gs) and hard (T_gh) segments that shifted to higher temperatures as the M_n decreased from 3k to 1k. This indicates that the increased content of hard segments restricted polymer chain mobility, thus reducing thermally induced dissociation of the cross-linked polymer networks (42, 65). Additional thermal characterizations using thermogravimetric analysis and differential scanning calorimetry are shown in fig. S10. Figure 2 (D and E) presents a set of images of the time-dependent recovery of SH-PLCLs at RT. The SH-PLCL films fully restored mechanical properties with ~500% stretchability after 12 hours (Fig. 2D) and even a short healing period of 1 min enabled to withstand bending and twisting motions (Fig. 2E). Analysis of the mechanical properties as a function of healing time in Fig. 2F exhibits that the recovery efficiency markedly increased to ~80% within 3 hours and reached ~100% at 12 hours. However, SH-PLCLs with low M_n values (1k and 2k) required elevated temperatures to promote the self-healing process, which originated from restricted disulfide metathesis due to inactive chain mobility below T_gh (Fig. 2G) (65). Other mechanical properties and self-healing behaviors appear in fig. S11. Considering potential applications to biological environments, we also evaluated the self-repairing efficiency of SH-PLCLs in water (Fig. 2H). SH-PLCL films were cut, reattached in water, and left to self-heal at various temperatures for 1 hour. Although the efficiencies in water were slightly different from those in air, the underwater efficiencies remained sufficiently high due to dynamic disulfide and hydrogen bonds (42, 43).

Fig. 2. Biochemical, mechanical, and self-healing characteristics of SH-PLCL elastomers.

(A) Time-dependent changes in weight ratios of SH-PLCL films with different molecular weights (M_n) of the PLCL prepolymer in PBS (pH 7) at RT. (B and C) Stress-strain curves (B) and DMAs (C) of SH-PLCLs as a function of M_n. (D) Sequential photographs representing self-healing of the SH-PLCL_3k film: as-prepared (top left), cut (top middle), self-healed (top right), and stretched (bottom). (E) Optical images of bent and twisted SH-PLCL_3k films after healing at RT for 1 min. (F) Stress-strain curves of pristine SH-PLCL (black) and others after being cut and self-healed for 1 hour (red), 3 hours (blue), and 12 hours (green) at RT. (G) Temperature-dependent recovery efficiencies of the mechanical properties of SH-PLCL films with various M_n values for a self-healing time of 12 hours. (H) Self-healing efficiency of SH-PLCL_3k films at various temperatures under air/water for 1 hour.

Stretchable SH-CCs

Figure 3A shows a set of images of freestanding SH-CC films that are capable of autonomously restoring electrical properties at RT after physical damages (bottom), achieved by the synergistic effect of hydrogen bond–rich PEG and glycerol. In Fig. 3B, high PEG ratios (>4) in the presence of glycerol accelerated the self-healing process and fully recovered electrical performance within ~2 s after repeated cuts (fig. S12). PEG also induced a phase separation of PEDOT and PSS domains (48), resulting in notable enhancement of conductivity (~400 S/cm), which was 20-fold higher than the one without the addition of PEG. Glycerol, which contains three hydroxyl groups, also increased conductivity but did not substantially influence the self-healing property compared to PEG. This is presumably due to its smaller molecular size and higher viscosity, which contributed to lower molecular mobility and limited interaction with other polymer chains compared to PEG (fig. S13). Consequently, through meticulous stoichiometric control, we selected PEG and glycerol ratios of 4 and 5 to achieve both high conductivity and effective self-healing. Figure 3C presents the influence of P14[TFSI] on electrical and mechanical properties of SH-CC. The result indicated that the ionic liquid played a role in improving the elasticity due to the mobility of PEDOT:PSS polymer chains (8, 35), while an additional quantity somewhat reduced the electrical conductivity. Such mechanical performance can be enhanced via formation of a thin SH-CC film (2 μm in thickness) through spin-coating onto an SH-PLCL substrate (100 μm in thickness) (Fig. 3D). The layered conductors provided electrical connections to light-emitting diodes (LEDs) and recovered the function even after complete cuts. Figure 3E displays continuous, real-time measurements as physical separation and self-healing repeatedly occurred through several cuts of the conductive composites. Although the healing time was slightly varied from that of the SH-CC films alone, the conductor consistently restored the original conductivity in ~100 s, outperforming physical reconstructions of previously reported self-healing conductors (37, 45). Furthermore, the self-repaired conductor exhibited outstanding electrical stretchability up to ~500%, although a gradual decrease in electrical conductivity was observed (Fig. 3F). Cyclic tests exhibited robust stability, retaining 80% of the initial conductance under 60% strain, due to the strong interface between the SH-CC film and SH-PLCL substrate (fig. S14).

Fig. 3. Stretchable, biodegradable, self-healing conductor based on SH-CC and SH-PLCL.

(A) Images of a freestanding SH-CC film (top) and SEM images of the cut and healed region (bottom). (B) Dependence of the conductivity and self-healing ability of SH-CC on the content of PEG. (C) Variation in the conductivity and stretchability of SH-CC in relation to the P14[TFSI] content. (D) Optical images of an SH-CC on SH-PLCL serving as an interconnection for an LED: original, cut, and self-healed for 1 min at RT. (E) Measured real-time currents as the recovery of electrical performance of SH-CC on SH-PLCL after repetitive sharp cuts. (F) Change in conductivity of SH-CC on SH-PLCL under tensile strains. Inset shows the stretched image of the film after self-healing. (G) Schematic (top) and optical images of an SH-CC interfacial layer sandwiched between top/bottom SH-PLCL layers before (middle) and after stretching (bottom), highlighting robust adhesion between layers of SH-PLCL and SH-CC. (H) Comparison of lap shear forces of glass, PDMS), PI, PLCL, and SH-PLCL stacks with an SH-CC interfacial layer. (I) Lap shear forces of PLCL and SH-PLCL stacks with different SH-CC interfacial layers such as PEDOT:PSS, PPy, and PANI. (J to L) Images (J), electrochemical impedance spectra (K), and cyclic voltammograms (L) of SH-CC electrodes patterned on SH-PLCL and PLCL substrates before and after sonication in deionized water.

The enhanced features of the combined composite were possible due to strong adhesion (67, 68), which was confirmed through lap shear tests [American society for testing and materials (ASTM) D3163] using a sandwich configuration of SH-CC and SH-PLCL as shown in Fig. 3G. We evaluated the lap shear force of SH-CC on diverse substrates including glass, polydimethylsiloxane (PDMS), polyimide (PI), PLCL, and SH-PLCL. The results revealed that the combined use of SH-PLCL achieved the highest shear strength (~160 kPa) among those substrates, which was attributed to interpenetrated polymer chains of SH-CC and SH-PLCL along with subsequent hydrogen bonding. Experimental validations including elemental analysis based on scanning electron microscope (SEM) image and chemical analysis appear in figs. S15 and S16. This approach was also effective in other conducting polymers such as polypyrrole (PPy) and polyaniline (PANI), resulting in substantially enhanced adhesion strengths with SH-PLCL compared to pristine PLCL (fig. S17). COMSOL simulation revealed the dependence of stretchability on adhesive force characterized by a dimensionless quantity, normalized adhesive stiffness, K_n, defined as (68)

$$K_n=\frac{\raisebox{1ex}{$\mathrm{\sigma }$}\!\left/ \!\raisebox{-1ex}{${\mathrm{\sigma }}_Y$}\right.}{\raisebox{1ex}{${\mathrm{\delta }}_n$}\!\left/ \!\raisebox{-1ex}{$h$}\right.}=\frac{\raisebox{1ex}{$\mathrm{\sigma }$}\!\left/ \!\raisebox{-1ex}{${\mathrm{\delta }}_n$}\right.}{\raisebox{1ex}{${\mathrm{\sigma }}_Y$}\!\left/ \!\raisebox{-1ex}{$h$}\right.}=\frac{K}{\raisebox{1ex}{${\mathrm{\sigma }}_Y$}\!\left/ \!\raisebox{-1ex}{$h$}\right.}$$ (1)

where K, σ_Y, σ, δ_n, and h are the adhesive stiffness, yield strength, interfacial strength, opening displacement, and thickness of the top conductive layer, respectively, with the normalized contact gap given by δ_n/h (Fig. 3H and figs. S18 and S19). Delamination (i.e., the opening of contact gaps) occurred for a weak interfacial adhesion strength between the conductive layer and elastomeric substrate (K_n, 0.001) upon stretching, resulting in stress localization and subsequent rupture. Conversely, an increase in interfacial adhesion reduced delamination or contact gaps, with no fracture observed at a high adhesion strength (K_n, 1). This inverse relationship between normalized adhesive stiffness and contact gap confirmed the high stretchability of the proposed conductor, with a K_n of ~2 (Fig. 3I and fig. S18). We note that an increase in the thickness (h) reduced adhesive stiffness for a given K_n, resulting in large contact gaps and mechanical failures under strains (Eq. 1 and fig. S20). In this respect, the thin geometry of the SH-CC is crucial for superior stretchability.

The strong adhesion can ensure both physical and electrical stabilities even under aqueous conditions. Figure 3J illustrates the self-healing conductor-based electrophysiological sensing electrodes on PLCL (top) or SH-PLCL (bottom) sheets before (left) and after (right) sonication in a PBS (pH 7) solution. Apparent damage or delamination was observed in the SH-CC on PLCL after sonication for 5 min, whereas the electrodes on the SH-PLCL remained intact even after sonication for 1 hour. Figure 3 (K and L) confirmed the electrochemical stability through measurements of impedance (Fig. 3K) and cyclic voltammetry (Fig. 3L) analyses. Both electrodes on the PLCL and SH-PLCL (original, black) showed nearly identical impedance (100 ohms at 1 kHz) and charge storage capacity (1.7 mC/cm²); however, these values of the devices on the PLCL (blue) were substantially deteriorated after sonication for 5 min. On the other hand, the parameters for the SH-PLCL (red) remained unchanged even after sonication for 1 hour, highlighting the potential applicability in wet, humid environments. Long-term stability tests of the electrode characteristics in PBS under physiological conditions (pH 7, 37°C) appear in fig. S21.

In vivo demonstration of self-healing electronics for urinary diseases

Careful examination of the self-recovery performance under environments such as the body or physiological conditions can provide an important tool for applicability in biomedical studies. We used the urinary bladder because the curvilinear, time-dynamic organ with a volume change of ~300% compared to static organs/tissues would have a high risk of destructions/failures and would be one of the harshest environments for self-recovery (7). Figure 4A illustrates a soft, implantable, self-healing electronic system that was designed to monitor and treat physiological functions of the bladder. This device consisted of SH-CC electrodes for recording electromyography (EMG) and delivering electrical stimulation as well as a capacitive strain gauge for monitoring bladder pressures; all of which were built on SH-PLCL substrates/encapsulants (Fig. 4B and fig. S22). The system featured two square pads at both ends of the substrate, allowing for suture-free integration onto the urinary bladder of a rat through the strong cohesion force (fig. S23), and the intrinsic softness and thin geometry (thickness, 100 μm) ensured minimization of the mechanical mismatch between the electronics and organ (Fig. 4C). Electrical and physiological studies using individual components appeared in Fig. 4 (D to F). The strain sensor exhibited a linear response to external strains up to 60%, similar to the radial expansion rate of the urinary bladder (volume change fraction, 300%) (Fig. 4D). A pair of electrodes was able to detect voiding events via EMG measurements (Fig. 4E) and induce urination via electrical stimulation, with a voiding efficiency comparable to that of normal urination (Fig. 4F and fig. S24). The overall evaluation of the self-healing property was performed by continuous, real-time observation of multiple parameters in the implanted electronic system (Fig. 4G). While monitoring the process of periodic micturition through intravesical pressures, strains, and EMG signals, we impaired interconnects of the strain and EMG sensors on purpose. Although the damages immediately resulted in signal disruptions, the system swiftly restored the original functions within 5 s, as shown in the magnified views.

Fig. 4. Self-healing, multifunctional medical implant for urinary diseases.

(A) Schematic representation of a self-healing E-web integrated onto the urinary bladder for monitoring and regulating the bladder’s biophysiological functions. (B) Optical images of the self-healing E-web featuring two sensing/stimulation electrodes and a strain gauge. Two patches located at both ends of the E-web enabled suture-free integration onto the bladder by abrading both surfaces with sandpaper and attaching them together via self-healing. (C) Photograph of the self-healing E-web implanted onto a rat urinary bladder. (D) Measured capacitance changes of the strain sensor as a function of strain. (E) EMG signals recorded during the bladder’s filling and voiding phases. (F) Comparison of bladder voiding efficiency during natural urination and electrical stimulation. (G) Representative traces of intravesical pressure, strain, and EMG signals, recorded using a pressure transducer and the self-healing E-web, during cystometry in response to repetitive bladder movements (left), with magnified views illustrating the recovery of electrical functions after intentional device scratches with a razor blade (right).

DISCUSSION

Synthesis, materials evaluations, and devices presented here describe a stretchable, biodegradable, highly conductive, and self-healing conductor for mechanically and electrically resilient electronic devices with diverse and sophisticated functionalities. The key constituents, i.e., SH-PLCL and SH-CC, exhibited autonomous self-healing capabilities, enabling the conductor to completely restore its exceptional conductivity (~1000 S/cm) within ~30 s at RT, which outperformed properties of previously reported self-healing conductors. Chemical interactions between SH-PLCL and SH-CC offered a sufficient adhesion strength of ~160 kPa, ensuring robust, reliable electrochemical performance even under harsh conditions. As a proof-of-concept demonstration, an integrated self-healing system was able to monitor and control the physiological functions of the urinary bladder in real time.

MATERIALS AND METHODS

Synthesis of SH-PLCL elastomers

The synthesis of SH-PLCL was based on our previous work with minor changes (8). ε-Caprolactone (ε-CL; Alfa Aesar United States; 10~40 mmol), l-lactide (LA; TCI, Japan; 10~40 mmol), stannous octoate [Sn(Oct)₂; Alfa Aesar, United States; 0.1 ~ 0.4 mmol], and butanediol (Alfa Aesar, United States; 5 mmol) were added into a dried three-necked flask equipped with a magnetic stirring bar. The amounts of ε-CL, LA, and Sn(Oct)₂ were varied according to the target molecular weight of PLCL-diol. The flask was sealed under vacuum after being purged three times with a N₂ atmosphere at 90°C, stirred in an oil bath at 150°C for 24 hours, and cooled to RT. IPDI (Sigma-Aldrich, United States; 10 mmol) dissolved in N,N′-dimethylformamide (DMF) was added dropwise into the flask and stirred at 60°C for 6 hours. The flask was cooled to RT again and added with bis(4-hydroxyphenyl)disulfide (TCI, Japan; 5 mmol) dissolved in DMF, followed by reaction at 70°C for 12 hours. The final product was precipitated three times in excess methanol and dissolved in DMF to achieve a concentration of 15% w/v. Last, an appropriate amount of the solution was poured into a PDMS mold and dried at 70°C for 12 hours to obtain an SH-PLCL film. For visualization of SH-PLCL, 0.1 mol % of the fluorescent dye, fluorescein isothiocyanate (Sigma-Aldrich, United States), was added to the solution.

Synthesis of SH-CCs

An SH-CC mixture was prepared by adding PEG 400 (Sigma-Aldrich, United States), glycerol (Sigma-Aldrich, United States), and P14[TFSI] (Sigma-Aldrich, United States) to a PEDOT:PSS (Clevios PH1000, Heraeus GmbH, Germany) solution in varying weight ratios of PEDOT:PSS (1), PEG (0 to 4), P14[TFSI] (0 to 3), and glycerol (5). The mixture was vigorously stirred using a magnetic stirrer at RT for 12 hours. Subsequently, the resulting solution was poured into a PDMS mold and allowed to dry at 50°C for 3 hours and at 140°C for 6 hours, yielding an SH-CC film. This two-step drying process ensured the formation of a uniform film without bubbles. Film thickness was measured using a DektakXT (Bruker, United States), and SEM images were captured with an S-4300 (Hitachi, Japan) to elucidate the self-healing behavior. For PPy- and PANI-based SH-CCs, PPy and PANI, synthesized following the previous literature (44), were used in place of PEDOT:PSS.

Fabrication of a stretchable, biodegradable, self-healing electronic tree

The electronic tree is composed of four components: top SH-CC, SH-PLCL interlayer, bottom SH-CC, and SH-PLCL substrate (fig. S1). Initially, an SH-PLCL film (100 μm) was prepared on a PDMS-casted glass substrate. Then, the SH-CC solution was spin-casted onto the SH-PLCL film using a PDMS stencil mask to create a pattern of humidity-temperature sensors and upper electrodes for a capacitive pressure sensor array, followed by drying on a hotplate at 50°C for 10 min. Following the mask removal, an annealing process at 140°C for 1 hour was conducted to ensure thorough drying and robust adhesion between the SH-CC (5 μm) and SH-PLCL layer. A similar procedure was applied to pattern another SH-CC layer onto an SH-PLCL film, forming bottom electrodes for the pressure sensor array. The pair of SH-CC/SH-PLCL layers was aligned and assembled together by applying heat (80°C) and pressure (~100 kPa) for 3 hours. After the establishment of electrical connections using a flexible anisotropic conductive film (ACF; Elform, United States), the device was gently detached from the PDMS substrate. Here, PDMS stencil masks (~200 μm) were fabricated using an ultraviolet laser cutter (MD-U1000C, Keyence, Japan).

Fabrication of a self-healing electronic web

The fabrication of the self-healing electronic web (E-web) followed a process similar to that of the electronic tree with minor adjustments. An SH-CC solution was spin-casted onto an SH-PLCL substrate (100 μm) using a PDMS stencil mask to achieve sensing/stimulation sensors and a lower electrode for a capacitive strain gauge and then dried on a hotplate at 50°C for 10 min. After removing the stencil mask, the SH-CC (5 μm)/SH-PLCL layers underwent annealing at 140°C for 1 hour. An SH-PLCL interlayer dielectric film, defined via laser cutting, was then assembled on top, and another SH-CC layer was patterned to complete the strain gauge. The fabrication process concluded with the integration of an SH-PLCL encapsulation layer and flexible ACF cables. The SH-PLCL substrate/encapsulation layers were patterned into a serpentine configuration using laser cutting.

In vivo experiments

All animal experiments were performed in accordance with the Korean Ministry of Food and Drug Safety Guide for the Care and Use of Laboratory Animals and were conducted according to protocols approved by the Institutional Animal Care and Use Committee (IACUC). This study was reviewed and approved by the IACUC of Samsung Biomedical Research Institute (SBRI) (approval number: SBRI-IACUC20230526001). SBRI is an Association for Assessment and Accreditation of Laboratory Animal Care International accredited facility and abide by the Institute of Laboratory Animal Resources guide. The animals were anesthetized by inhaling isoflurane (concentrations ranging from 3 to 5%; O₂ level, 0.1 liters min⁻¹) and remained under anesthesia throughout the bladder catheterization procedure and device implantation. The bladder was exposed through a midline incision in the abdomen, and a catheter was inserted via the bladder dome. The opposite end of the catheter was connected to a pump for saline infusion and a pressure transducer to monitor the bladder pressure. Integration of the self-healing E-web onto the bladder was then performed. Concurrently, evaluations of physiological parameters using the self-healing E-web, cystometrograms, and weighing of excreted urine were simultaneously conducted while saline was instilled at a rate of 1 ml hour⁻¹. Intentional physical damages on the interconnects for the strain and EMG sensors were conducted to evaluate self-healing capabilities of the E-web. Inductance, capacitance, and resistance (LCR) meter (IM 3533-10, HIOKI, Japan) and a customized program (LabVIEW, National Instruments) recorded the real-time capacitance of the strain gauge. EMG signals were amplified by a factor of 1000 using a differential ac amplifier (model 1700, A-M Systems, United States) and sampled at a frequency of 1 kHz via a data acquisition (DAQ) system (National Intrsuments, United States). Simultaneously, a pressure transducer in cystometry also monitored the intravesical pressure at a sampling rate of 200 Hz via the LabChart Reader software (AD Instruments, Australia). An electrical stimulator (model 2100, A-M Systems, United States) delivered biphasic pulses (2 mA, 5 ms, and 50 Hz) to the urinary bladder, promptly inducing urination.

Supplementary Materials

This PDF file includes:

Supplementary text

Figs. S1 to S24

Tables S1 and S2