Electron-ion coupling enables ionic hydrogel with high thermopower for low-grade heat harvest and sensitive fire warning

Abstract

Conventional ionic thermoelectric materials suffer from inherent intermittent power generation due to their reliance on ionic thermal diffusion. To overcome this limitation, we developed a flame-retardant thermoelectric hydrogel leveraging electron-ion thermoelectric synergy. The hydrogel was prepared through thermal polymerization using sodium alginate (SA), poly(3,4-ethylenedioxythiophene):polystyrene sulfonate (PEDOT:PSS), polyacrylamide (PAAm), and sodium chloride (NaCl). Subsequent immersion in calcium chloride (CaCl₂) solution enabled strong SA-Ca²⁺ crosslinking, enhancing the flame retardancy and mechanical strength to yield a high-performance ionic thermoelectric hydrogel (APSG). The synergy between the electronic Seebeck effect (PEDOT:PSS) and ionic thermal diffusion (NaCl) achieved a high Seebeck coefficient of 4.25 mV K⁻¹. Under a 20 K temperature difference with voltage amplification, APSG generated a sustained output voltage of 2.5 V for over 10 min, demonstrating robust continuous power generation. APSG also exhibited sensitive fire-warning capability and outstanding flame retardancy. When exposed to a flame, it triggered the fire alarm within 1.7 s. Its limiting oxygen index (LOI) was as high as 42.3%, and achieved a UL-94 V-0 rating. This work presents an effective strategy for developing next-generation safe and efficient thermoelectric materials for energy harvesting.

---

1 Introduction

The unprecedented surge in global energy demand since the early 21st century has driven excessive fossil fuel consumption, exacerbating energy crises and accelerating ecological degradation.^1,2 This urgent challenge has galvanized extensive research into energy harvesting and recycling technologies, among which low-grade heat recovery represents a pivotal strategy for sustainable energy utilization.^3–6 Thermoelectric (TE) materials, capable of directly converting thermal energy into electrical energy,^7–9 hold significant promise for energy applications due to their high conversion efficiency, environmental compatibility, and self-powering nature.^10–13

TE materials are broadly categorized by charge carriers as electronic-type (e-TE) or ionic-type (i-TE).^14,15 While e-TEs exhibit high electrical conductivity suitable for direct circuit integration, their intrinsically low Seebeck coefficients (<200 μV K⁻¹) necessitate complex p–n junction configurations.^16,17 In contrast, i-TEs leverage thermal diffusion/current phenomena^18–20 to generate thermovoltage 2–3 orders of magnitude higher under minimal temperature gradients, attributable to asymmetric cationic/anionic migration kinetics.^21–24 Notably, solid/quasi-solid i-TEs (e.g., ionic thermogel polymer electrolytes, i-TGPEs) offer unique flexibility, biocompatibility, and multifunctionality—including flame retardancy, self-healing, and adhesion.^25–28 Despite these advantages, i-TGPEs still have critical limitations.

Conventional i-TGPEs exhibit substantially lower electrical conductivity than e-TE materials. Their energy conversion fundamentally relies on ionic thermal diffusion, wherein charged ions cannot directly participate in external circuit conduction. Consequently, power output requires intermittent charge–discharge cycles mediated by electric double-layer capacitance at electrode interfaces,²⁹ imposing severe constraints on practical applications. Furthermore, the inherent flammability of organic polymer matrices raises significant safety concerns.³⁰ These dual challenges underscore the critical need for developing next-generation intelligent flame-retardant i-TGPEs featuring innovative thermoelectric conversion mechanisms.

To address these issues, we have developed a novel electron-ion coupled multifunctional flame-retardant ionic hydrogel (APSG) via an interpenetrating polymer network strategy. This strategy integrated electronic-type thermoelectric polymers with ionic conduction channels, enabling a unique ion-electron thermoelectric synergy mechanism. Under a temperature gradient, the thermal diffusion of ions generated a built-in electric field that drove the directional drift of electrons/holes in the conduction band, effectively converting the ionic current into a sustainably output electron current without the need for capacitive intermediaries. This architecture of APSG featured in situ polymerized ionic gels embedded within a continuous electronic framework, establishing uninterrupted dual-conduction pathways for both ions and electron.³¹ The synergistic coupling of Seebeck effect and ionic thermal diffusion enabled APSG to deliver a stable 2.5 V output over 10 min under a minimal 20 K temperature gradient, and its Seebeck coefficient of 4.25 mV K⁻¹ exceeded typical ionic thermogels (≤1 mV K⁻¹) by over 400%, fundamentally overcoming the intermittent power supply limitation of conventional i-TEs (typically <30 s discharge). Remarkably, unlike flammable PVA-based systems requiring extrinsic fireproof coatings,³² APSG simultaneously achieved superior flame retardancy with sensitive fire-warning response, providing new insights for developing intelligent fire-safety thermoelectric systems.

2 Experimental section

2.1 Materials

The Materials are listed in the SI.

2.2 Preparation of APSG

Configuration schematic of ionic thermoelectric system is provided in Fig. S2, and APSG was synthesized through the following procedure (Fig. 1). First, a homogeneous dark-blue solution was prepared by dissolving poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS) aqueous dispersion and NaCl in 8 mL deionized water, followed by adding 0.2 g sodium alginate (SA) to achieve a 25 mg mL⁻¹ SA solution under 12 h magnetic stirring at room temperature. Subsequently, 2 g acrylamide (AAm) and 1.5 mL deionized water (H₂O) were incorporated with continuous stirring until complete dissolution. The precursor solution was then mixed with 0.3 mL ammonium persulfate (APS, 1.5 wt% relative to AAm monomer), 0.2 mL N,N′-methylenebisacrylamide (MBA) crosslinker and 20 μL N,N,N′,N′-tetramethyle-thylenediamine (TEMED) redox catalyst. After 30 s vigorous stirring, the mixture was cast into polytetrafluoroethylene (PTFE) molds and polymerized at 65 °C for 2 h in a forced-air oven. The obtained hydrogel was immersed in 10 wt% CaCl₂ solution for 5 min to yield the PAAm-PEDOT:PSS-SA-NaCl composite hydrogel (APSG). The detailed formulation provided in Table. S1.

Fig. 1 Schematic diagram of fabrication of high-efficiency thermoelectric flame-retardant ionic hydrogel (APSG) via electron-ion synergistic coupling.

2.3 Preparation of Wood@APSG

The as-prepared APSG hydrogel (0.5 mm thickness) was cut to size and directly adhered to Scotch pine wood substrates (Wood) using its intrinsic self-adhesive property, forming a flame-retardant composite material (Wood@APSG) with real-time fire-warning functionality.

2.4 Preparation of thermoelectric supercapacitors based on APSG

As shown in Fig. S1, the APSG thermoelectric supercapacitor (APSG SC) was fabricated as follows. A pre-synthesized APSG hydrogel was precision-cut into 10 mm × 10 mm × 1.5 mm slices. Subsequently, composite carbon cloth electrodes consisting of activated carbon, conductive carbon black, and polyvinylidene fluoride (mass ratio 8 : 1 : 1) with N-methylpyrrolidone as solvent were attached to both of the hydrogel surfaces. To establish conductive pathways, copper foil current collectors (30 mm × 10 mm × 0.1 mm) were then laminated onto the electrodes, maintaining an effective contact area of 10 mm × 10 mm between the foil and carbon cloth. Finally, prior to encapsulation with polyimide tape, silicone buffer layers (10 mm × 10 mm × 2 mm) were incorporated at both termini to complete the fabrication of APSG SC.

2.5 Characterization

The Characterization are listed in the SI.

3 Results and discussion

3.1 The morphology and mechanical properties of APSG

Fig. 2 illustrates the morphology and mechanical properties of APSG hydrogels. With increasing PEDOT:PSS content, the hydrogel transitioned progressively from a transparent gel to a dark blue-black state (Fig. 2a), confirming the successful incorporation of PEDOT:PSS. Further observations were performed through Scanning electron microscopy and energy dispersive X-ray spectrometry (SEM-EDS). Cross-sectional SEM images (Fig. 2b) revealed uniformly distributed micro/nano-scale porous channels throughout the hydrogel, along with homogeneous dispersion of flame-retardant elements such as N, S, Cl, and Ca (Fig. 2c), further confirming the successful fabrication and multifunctional integration of the APSG structure. FTIR analysis (Fig. 2d) identified the characteristic peaks for PAAm and SA, including the N–H and C=O stretching vibrations at 3351 cm⁻¹ and 1673 cm⁻¹, respectively, as well as the asymmetric C–H and COO⁻ stretching vibrations at 2929 cm⁻¹ and 1606 cm⁻¹.^33,34 These peaks exhibited slight redshifts at higher PEDOT:PSS content, suggesting enhanced hydrogen bonding and electrostatic interactions among PEDOT:PSS, PAAm, and SA networks, thereby promoting formation of an entangled triple-network structure. Furthermore, the intensity reduction or disappearance of several characteristic peaks of AAm, SA, and PEDOT:PSS implies not only possible attenuation due to reduced functional group concentrations or solvent shielding, but also provides additional evidence of PAAm polymerization and deep interpenetration of the three-component polymer networks.^35,36

Fig. 2 (a) Digital photographs of APSG with varying PEDOT:PSS content and video snapshots illustrating their flexibility under bending, twisting, and knotting; the SEM (b) and EDS (c) images of APSG; (d) FTIR spectra of AAm, SA, PEDOT:PSS, APSG2 and APSG4; (e) stress–strain curves of APSG0-APSG6; (f) comparison of tensile strength and fracture elongation; (g) cyclic stress–strain curves of APSG4 under 200% strain for 30 loading–unloading cycles.

Benefiting from the synergistic effect of cross-linking and molecular entanglement,³⁷ the APSG hydrogels exhibited good mechanical robustness, sustaining 180° bending, twisting, and knotting deformations (Fig. 2a), as well as displaying outstanding tensile/compressive resilience (Fig. S3), fully satisfying the mechanical requirements for solid-state flexible electrolytes. Tensile tests (Fig. 2e–g) revealed that all the APSG samples exhibited tensile strengths exceeding 1.20 MPa and elongations at break above 700%. Notably, the tensile strength increased from 1.24 MPa (APSG0) to 1.49 MPa (APSG4) with increasing PEDOT:PSS content, attributed to the formation of a dual-network structure where PEDOT:PSS entangled with the CA gel network, enhancing the structural support and mechanical strength. As shown in Fig. 2f, the stress–strain curves of APSG4 under 200% strain remained nearly unchanged after 30 successive loading–unloading cycles, confirming its excellent flexibility and cycling durability.

3.2 Thermoelectric performance and thermoelectric capacitance performance

Within the PAAm-CA-PEDOT:PSS molecular network, hydrogen bonding and electrostatic interactions synergistically enhanced the mechanical strength of APSG while concurrently establishing efficient ion transport channels and reinforcing the PEDOT:PSS electronic backbone. This structural integration transformed the initially loose conductive polymer pathways into a more compact and robust configuration, providing a solid foundation for high i-TE performance. Compared to the control sample APSG0 and APSG5 without NaCl, the other APSG formulations exhibited a distinct linear correlation between thermovoltage and temperature gradient (Fig. 3a and b). The introduction of NaCl supplied abundant free Na⁺ and Cl⁻ ions, enabling effective ionic thermodiffusion within the hydrogel matrix. Furthermore, the abundant amide (–CO–NH₂) and carboxyl (–COO⁻) groups in the PAAm and CA chains established favorable interactions with the sulfonate moieties of PSS in PEDOT:PSS, which created efficient Na⁺ migration pathways for Na⁺ migration, thereby amplifying the migration disparity between Na⁺ and Cl⁻ ions and significantly enhancing the ionic thermopower. Consequently, the Seebeck coefficient increased from nearly 0 in APSG0 to 3.65 mV K⁻¹ in APSG5, indicating a p-type i-TE behaviour. Increasing the PEDOT:PSS content further elevated the Seebeck coefficient from 3.74 mV K⁻¹ to 4.25 mV K⁻¹, which was attributed to the electronic thermoelectric contribution of the PEDOT:PSS network.³⁸ Under a temperature gradient, charge carriers (holes) within the PEDOT:PSS backbone underwent directional migration toward the cold end, analogous to ionic thermodiffusion, thereby augmenting the thermovoltage. Moreover, the internal electric field generated by Na⁺ and Cl⁻ thermodiffusion further promoted charge carrier movement, which enabled the effective conversion of ionic thermodiffusion into electronic current output,^35,39,40 allowing APSG to generate continuous external power. As illustrated in Fig. 3b, c and S4, APSG samples devoid of Na⁺ and Cl⁻ ions exhibited markedly diminished thermoelectric performance (Seebeck coefficient = 0.35 mV K⁻¹), approaching values characteristic of typical electronic thermoelectric materials.³⁸ These results demonstrated that the thermoelectric conversion mechanism of APSG fundamentally differed from conventional single-mode thermoelectric systems, arising from a synergistic interplay between ionic and electronic thermoelectric effects.

Fig. 3 Thermoelectric performance of APSG. (a) Voltage response curve of APSG4; (b) linear fitting curves of thermal voltage vs. temperature difference and (c) corresponding Seebeck coefficients of APSG0-APSG6; (d) schematic illustration of the thermocapacitor working mechanism; (e) eeal-time output voltage profiles during thermoelectric conversion; (f) current–voltage characteristics and (g) power–voltage output curves of APSG thermocapacitor under different temperature gradient.

As illustrated in Fig. 4, the coupled electron-ion thermoelectric mechanism in APSG operated through two parallel pathways: (i) directional migration and accumulation of holes/electrons in the PEDOT:PSS backbone at the cold end under a thermal gradient, and (ii) thermodiffusion of Na⁺ and Cl⁻ ions from the hot to cold side. The smaller radius, higher mobility, and stronger electrostatic interaction of Na⁺ with the polymer matrix drove pronounced ion migration asymmetry, resulting in substantial ionic Seebeck effect. Additionally, the induced electric field generated by ionic thermodiffusion further accelerated carrier migration, thereby amplifying the ionic thermoelectric current. Consequently, APSG achieved synergistic thermoelectric conversion via dual electronic and ionic transport channels. This strategy effectively combined the advantages of conventional electronic and ionic thermoelectric systems while overcoming their individual limitations (e.g., low voltage output in ionic systems and limited current in electronic systems), significantly broadening its potential applications. Benefiting from its unique electron-ion coupled thermoelectric conversion mechanism, APSG exhibited significant potential for next-generation energy storage and conversion devices. As shown in Fig. 3d, the thermocapacitor based on APSG operated through a four-stage cycle mechanism,^26,41 with real-time voltage response depicted in Fig. 3e. In stage I, applying a temperature gradient induced directional migration of free Na⁺ and Cl⁻ ions within APSG. Their differential mobilities resulted in net charge accumulation at the cold end, generating a positive thermoelectric potential (higher voltage at the cold side). During stage II, external load connection promoted electric double-layer formation at the APSG-electrode interface. Electrons from the hot-side electrode flowed through the external circuit to the cold side, establishing directional current and gradually electrostatic equilibrium. While thermovoltage decreased during this phase, sustained power output was maintained. In stage III, load disconnection and temperature gradient removal allowed ions to return to equilibrium distribution. However, accumulated charges gathered on the electrodes generated a reverse-polarity thermovoltage across the device. Finally, in stage IV, load reconnection discharged residual charges through the external circuit, enabling secondary power output while resetting the system to its initial electrostatic state (thermovoltage = 0). This cyclic process enabled continuous thermal-to-electric energy conversion. As shown in Fig. 3f–g, under ΔT = 3, 6, 9, and 12 K, the thermocapacitor achieved open-circuit voltage of 12.8 mV, 25.6 mV, 37.7 mV, and 51.2 mV, respectively, with corresponding short-circuit current of 2.4 μA, 6.4 μA, 9.7 μA, and 11.8 μA. The maximum output power escalated from 7.6 nW to 151.5 nW, showing clear temperature-dependent enhancement. These results indicated that the APSG thermocapacitor could efficiently harvest and store low-grade thermal energy at modest temperature differences, exhibiting exceptional thermocapacitive behaviour and reliable energy output. The linear thermoelectric response to applied ΔT validated APSG's practical viability for thermal management and low-grade energy harvesting applications beyond conventional thermoelectric materials.⁴²

Fig. 4 Schematic diagram of electron-ion coupled thermoelectric conversion mechanism of APSG.

3.3 Fire-warning capability

The unique electron-ion coupled thermoelectric conversion mechanism endowed APSG with exceptional fire warning sensitivity. As illustrated in Fig. 5a, the fire alarm devices fabricated by APSG2 and APSG4 (both incorporating NaCl as the ion source and PEDOT:PSS as the electron conductor) triggered reliable fire warnings with excellent repeatability. Upon flame exposure, APSG2 activated the fire alarm within ∼4.0 s (Fig. 5b and S5), while the higher PEDOT:PSS content in APSG4 enabled a significantly faster response—generating 80.3 mV thermoelectric voltage in just about 1.0 s. Removal of the flame restored the output voltage of both samples to baseline, confirming good thermoelectric reversibility. Repeated ignition tests further demonstrated that both APSG2 and APSG4 could consistently re-trigger the alarm (Fig. 5b–d), indicating their stable and reliable fire sensing behaviour.⁴³ As shown in Fig. 5e, compared with APSG2, APSG4 exhibited superior sensitivity, reducing fire warning time from ∼5.0 s to 2.0 s. Furthermore, Fig. 5f revealed that increasing the PEDOT:PSS content at fixed NaCl concentration a markedly decreased response time from 5.3 s (APSG1) to 1.7 s (APSG4). This enhancement arose from the synergistic electron-ion thermoelectric mechanism: under temperature gradients, Na⁺ and Cl⁻ ions underwent directional migration to generate ionic thermopower, while holes/electrons in the PEDOT:PSS network also migrated under the induced electric field, producing a usable electronic current through an external circuit. This coupling accelerated charge accumulation at the electrodes, enabling rapid and precise fire warning detection. In contrast, APSG5 (solely ionic thermodiffusion) required ∼5.4 s to trigger the alarm, while APSG6 (purely electronic thermoelectricity) showed the slowest response (∼8.0 s). These results further verified the pivotal role of electron-ion synergy in enhancing both thermoelectric performance and fire warning sensitivity in APSG.

Fig. 5 Fire warning performance of APSG. (a) schematic diagram of the fire alarm test setup; (b) video screenshots of fire alarm tests for APSG0 and APSG4; (c and d) fire-induced voltage response curves and (e) average fire warning trigger time of APSG2 and APSG4; (f) fire response time of APSG0-APSG6.

3.4 External energy supply application

Owing to its exceptional mechanical robustness and high conductivity, APSG could be utilized as a gel electrolyte for flexible energy storage and conversion devices. As shown in Fig. 6a and b, the electrochemical impedance spectroscopy (EIS) and ionic conductivity of APSG0-APSG6 were systematically investigated. While APSG0 exhibited high impedance (678.2 Ω) and low conductivity (0.02 S m⁻¹), APSG5 and APSG6 (modified with NaCl or PEDOT:PSS, respectively) demonstrated significantly reduced impedance (192.7 Ω and 372.2 Ω), and enhanced conductivities (0.08 S m⁻¹ and 0.04 S m⁻¹). Notably, increasing PEDOT:PSS content further reduced impedance while markedly improving conductivity: APSG4 achieved an impedance of 51.7 Ω and a conductivity of 0.29 S m⁻¹. This enhancement was mainly attributed to the formation of a continuous and efficient conductive network by PEDOT:PSS within the gel matrix. When integrated into an external power supply circuit, APSG was capable of forming a conductive path to illuminate a 2.5 V-rated light bulb (Fig. 6c and d). Notably, under various mechanical deformations such as stretching, 180° bending, and twisting, the circuit remained operational, confirming the uniformly distribution and strong interfacial boning of the PEDOT:PSS conductive framework. Furthermore, APSG spontaneously converted thermal energy to electrical energy through the coupled electron-ion thermoelectric mechanism under a temperature gradient, generating sustained thermoelectric voltage and current. As illustrated in Fig. 6d, a thermoelectric output system was assembled based on the ionothermal test platform, supplemented with a voltage amplifier and a load light bulb. Under a temperature difference of 20 K, APSG generated a stable thermovoltage of approximately 85.0 mV. After amplification by a factor of 30, the output voltage reaches ∼2.5 V, successfully powering the light bulb for over 10 min, highlighting its sustained energy supply capability.

Fig. 6 The external energy supply application of APSG. (a) EIS curves of APSG0-APSG6; (b) impedance values and ionic conductivity of APSG0-APSG6; (c) schematic diagram of the conducting pathway of APSG4 and digital photographs of the pathway under different deformations; (d) schematic diagram of the ionic thermoelectric power supply circuit of APSG; (e) digital photo of the ionic thermoelectric power supply test of APSG4 and (f) real-time voltage profile and voltage amplification profile.

Unlike the conventional ionic thermoelectric materials that rely on intermittent four-stage capacitive charging, APSG achieved continuous energy output through its inherent electron-ion coupling mechanism, offering a novel approach for thermoelectric conversion and overcoming limitations of traditional systems such as discontinuity, operational complexity, and limited applicability.

3.5 Flame retardancy and mechanism

Fire safety of electrolytes is crucial in practical applications. The thermogravimetric (TG) and derivative thermogravimetric (DTG) curves (Fig. S6a and b) of APSG0, APSG2, and APSG4 under ambient air atmosphere demonstrated good char-forming capability, with residual weights of 42.3 wt%, 42.9 wt%, and 45.5 wt%, and peak weight loss rates of 11.3 wt% min⁻¹, 9.9 wt% min⁻¹, and 10.3 wt% min⁻¹, respectively, which was mainly attributed to the efficient carbonization capability of the polymer network at elevated temperature.

As displayed in Fig. 7a and b, vertical combustion tests further demonstrated superior flame retardancy: both APSG0 and APSG4 rapidly self-extinguished after fire removal. Post-combustion structural integrity was maintained, with carbonization localized to the flame-contact region. To decouple the role of water evaporation, upon drying at 80 °C for 24 h in a blast oven, the dry gels of APSG0 and APSG4 were subjected to the vertical combustion test. The results demonstrated that the polymer networks remained structural stable after 10 s of flame exposure (Fig. 7c), forming dense char layer that effectively insulated heat and oxygen, confirming sustained flame retardancy without water's endothermic contribution.⁴⁴ Notably, APSG2 and APSG4 exhibited enhanced thermal stability and flame retardancy with higher PEDOT:PSS content. This improvement might arise from the incorporation of sulfur (S), an intrinsic flame-retardant element in PEDOT:PSS, and the formation of thermally stable calcium sulfate (CaSO₄) through high-temperature oxidation of Ca²⁺ and sulfonate groups.^33,45

Fig. 7 Screenshots of vertical combustion tests of (a) APSG0, (b) APSG4 and (c) APSG after drying; (d) SEM images of surface char layer (d₁) and elemental distribution of N, S, Cl and Ca in the char layer of APSG4 (d₂).

The microstructure and chemical composition of the post-combustion char layers were further characterized by SEM-EDS (Fig. 7d) and XRD (Fig. S6c). The XRD patterns of APSG4's char residue revealed prominent diffraction peaks at 2θ = 1.7°, 41.9°, 45.4°, 56.5° and 83.9°, corresponding to the (102), (211), (103), (311), and (314) planes of CaSO₄, in agreement with the standard JCPDS card no. 37-1496,⁴⁶ confirming the formation of a crystalline, thermally stable char structure. The SEM analysis revealed that the char layers formed by APSG after flame treatment were highly compact, providing effective barriers to heat and oxygen. The EDS mapping further indicated uniform distribution of flame-retardant elements such as S, N, and Ca in the char of APSG4, suggesting that the dense char layer synergistically formed by multiple flame-retardant elements in the condensed phase acted as an isolation effect and avoided further heat penetration into the internal matrix.⁴⁷

To further investigate the flame-retardant performance and underlying mechanism of APSG, the micro-scale combustion calorimetry (MCC) and limiting oxygen index (LOI) tests were conducted on APSG0, APSG2, and APSG4, as summarized in Table. S2. Due to the condensed phase flame-retardant effect from CA, APSG0 exhibited a high LOI value (39.5%) and achieved a UL-94 V-0 rating. With increasing the PEDOT:PSS content, APSG4 demonstrated significantly reduced flammability, evidenced by the substantial decrease in heat release capacity (HRC), peak heat release rate (PHRR), and total heat release (THR), from 288 J g⁻¹ K⁻¹, 239 W g⁻¹, and 16 kJ g⁻¹ to 263 J g⁻¹ K⁻¹, 224 W g⁻¹, and 14 kJ g⁻¹, respectively. Additionally, the PHRR peak shifted to a higher temperature of 286 °C, indicating enhanced thermal stability, while the LOI value of APSG4 further increased to 42.3%, confirming improved flame retardancy.

The thermogravimetric-Fourier transform infrared spectroscopy (TG-FTIR) analysis revealed that APSG4 released markedly less CO₂ during thermal decomposition. Moreover, the emission of flammable gases was notably reduced, whereas non-flammable species such as H₂O were generated in higher amounts (Fig. 8a–c and S7). These results suggested that the incorporation of PEDOT:PSS not only facilitated an electron-ion coupled system but also enhanced the thermal resistance, thereby retarding decomposition at elevated temperatures.⁴⁸ Furthermore, it could be indicated that APSG played a dilution effect through the heat-absorbing vaporization of water and the inert gases such as NH₃ released by decomposition, which in turn achieved the favorable flame-retardant effect in the gaseous phase.⁴⁹

Fig. 8 Investigation of the flame-retardant mechanism of APSG. (a) 3D IR spectra of APSG4 in air atmosphere; the corresponding FTIR spectra at different temperatures of (b) APSG0 and (c) APSG4; (d) XPS spectra of char layer of APSG4 after combustion and (e and f) C 1s spectra of APSG4 before and after combustion.

Elemental analysis of the post-combustion residues (Fig. 8d and S8) confirmed the presence of flame-retardant elements such as N, Cl, S, and Ca in both APSG0 and APSG4, which served to reinforce the flame-retardant capacity. The XPS analysis of the C 1s spectra before and after combustion (Fig. 8e and f) revealed that the proportion of C–C/C=C bonds in APSG4 increased from 53.8% to 60.5%, and the ester bond –O–C=O content rose from 1.6% to 2.7%, while the content of unstable C=O groups decreased from 7.5% to 5.3%. These changes indicated the formation of a more conjugated and compact char layer, which effectively enhanced thermal insulation and oxygen barrier properties. Furthermore, the C–S and C–N contents decreased from 18.3% and 15.6% to 14.6% and 8.2%, respectively, which further confirmed the formation of thermally stable CaSO₄via the high-temperature reaction between Ca²⁺ and sulfonate groups, along with the evolution of non-flammable gases such as NH₃, contributing to gaseous-phase flame retardancy.⁵⁰

In summary, the outstanding flame retardancy of APSG arose from the synergy of condensed-phase protection via a dense char layer and gaseous-phase dilution through the release of non-flammable species. The SA-Ca²⁺ crosslinked network and the PEDOT:PSS derived sulfur component synergistically catalyze the dehydration and carbonization, resulting in the formation of a dense char layer that acts as an effective heat/oxygen barrier to inhibit the release of flammable volatiles; The dilution effect combines the heat-absorbing vaporization of water and the release of inert gases such as NH₃ from decomposition to achieve self-extinguishing by simultaneously lowering temperature and combustibles availability. This dual mechanism significantly improved the thermal stability and fire safety of APSG, making it a promising ionic thermoelectric gel electrolyte candidate for applications in high-temperature or extreme environments.

4 Conclusions

In this work, we developed APSG, a novel ionic hydrogel featuring exceptionally high thermopower for low-grade heat harvesting and sensitive fire warning, via an interpenetrating polymer network strategy. This approach integrated electronic-type thermoelectric polymers with ionic conduction channels, resulting in in situ polymerized ionic gels within continuous electronic frameworks that ensured uninterrupted dual-conduction pathways. The synergistic coupling of the Seebeck effect and ionic thermal diffusion enabled APSG to deliver a stable 2.5 V output under a minimal 20 K temperature gradient, thereby overcoming the intermittent power supply limitation inherent in conventional i-TE materials. Remarkably, APSG concurrently achieved superior flame retardancy and rapid fire-warning response, triggering a fire alarm within 1.7 s of flame exposure. The hydrogel exhibited a limiting oxygen index of 42.3% and achieved a UL-94 V-0 rating. This study demonstrates an effective strategy for designing next-generation safe and efficient thermoelectric materials for energy harvesting applications. Further studies should explore to broaden the operational temperature window (e.g., via amphiphilic network designs) and enhance long-term hydration stability through covalent-coordination dual crosslinking strategies. Additionally, concurrent optimization of electrode interfaces could significantly boost energy conversion efficiency, making APSG more suitable for applications in extreme-environment energy harvesters and industrial fire-safety systems.

Data availability

The data that support the findings of this study are available from the corresponding author, Xuejun Lai, upon reasonable request.

Supplementary experimental procedure: The Experimental section details the Materials used (including sources and purity), the Preparation process for APSG and APSG SC (with specific formulations provided in Table S1), and comprehensive Characterization methods. These methods encompass Fourier transform infrared spectroscopy (FTIR) for chemical analysis, scanning electron microscopy and energy dispersive X-ray spectrometry (SEM-EDS) for morphology and elemental mapping, mechanical performance testing via uniaxial tensile tests, flame retardancy evaluation through limiting oxygen index (LOI), vertical burning (UL-94), and micro-scale combustion calorimetry (MCC), thermoelectric performance measurement using a custom setup (Fig. S2), fire-warning capability tests monitored with a millivolt alarm (Fig. S3), X-ray diffraction (XRD) for residue analysis, and electrochemical performance assessment (CV, GCD, EIS). Subsequent sections present results: Morphology and mechanical properties are illustrated in Fig. S3 (showing tensile/compressive deformation), Thermoelectric properties are displayed in Fig. S4 (voltage response curves), Fire-warning capability is demonstrated in Fig. S5 (test screenshots), and Flame retardancy and mechanism are analyzed using Fig. S6 (TG/DTG curves and char XRD), Table S2 (key flame-retardant parameters like HR capacity, PHRR, THR, LOI, and UL-94 rating), Fig. S7 (3D IR spectra), and Fig. S8 (char XPS spectra). Graphical schematics (Fig. S1, S2) and experimental data plots add the methodology and findings throughout. See DOI: https://doi.org/10.1039/d5ta05571g.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

We greatly acknowledge the National Natural Science Foundation of China (22478140, 22178134), and Guangdong Basic and Applied Basic Research Foundation (2024A1515010340, 2023A1515010845).

References

1. X. Chen, Y. Lin, B. Chen, R. Duan, Z. Zhou and C. Lu, Small, 2025, 21, 2412336.
2. B. Mohan, K. Singh, R. K. Gupta, A. J. L. Pombeiro and P. Ren, Prog. Mater. Sci., 2025, 152, 101457.
3. J. Liu, W. Jiang, S. Zhuo, Y. Rong, Y. Li, H. Lu, J. Hu, X. Wang, W. Chen, L. Liao, M. Zhuo and K. Zhang, Sci. Adv., 2025, 11, eadr2158.
4. Q. M. Saqib, M. Yousuf, M. Noman, A. Mannan, C. S. Patil, J. Kim, S. R. Patil, Y. Ko, N. R. Chodankar and J. Bae, Nano Energy, 2025, 133, 110456.
5. X. Hao, J. Wang and H. Wang, Chem. Soc. Rev., 2025, 54, 1957–1985.
6. X. Yang, P. Wang, X. Tang, Z. Wang, J. Ye, W. Duan, Y. Yue, T. Ci, Y. Liu and Y. Ju, Nano Energy, 2025, 140, 111057.
7. T. Rodrigues-Marinho, V. Correia, C.-R. Tubio, A. Ares-Pernas, M.-J. Abad, S. Lanceros-Méndez and P. Costa, Chem. Eng. J., 2023, 473, 145297.
8. S. Liu, D. Wu, M. Kong, W. Wang, L. Xie and J. He, ACS Energy Lett., 2025, 10, 925–934.
9. Z. Yu, X. Qu, M. Zhou, M. H. Mia, L. Hou, Y. Wang, J. Xu, L. Huang, L. Zhang, Q. Zeng and H. He, Chem. Eng. J., 2025, 515, 163622.
10. J. Ding, Y. Yang, J. Poisson, Y. He, H. Zhang, Y. Zhang, Y. Bao, S. Chen, Y. M. Chen and K. Zhang, ACS Energy Lett., 2024, 9, 1803–1825.
11. T. Li, X. Zhang, S. D. Lacey, R. Mi, X. Zhao, F. Jiang, J. Song, Z. Liu, G. Chen, J. Dai, Y. Yao, S. Das, R. Yang, R. M. Briber and L. Hu, Nat. Mater., 2019, 18, 608–613.
12. C.-G. Han, X. Qian, Q. Li, B. Deng, Y. Zhu, Z. Han, W. Zhang, W. Wang, S.-P. Feng, G. Chen and W. Liu, Science, 2020, 368, 1091–1098.
13. X. Xu, B. Yuan, Y. Chen, Z. Chen, B. Mao and P. Xiao, Chem. Eng. J., 2025, 513, 162924.
14. L. Cao, T. Sun, H. Zhao, L. Wang and W. Jiang, Chem. Eng. J., 2025, 506, 160206.
15. N. Fan, W. You, W. Gong, F. Jiao and W. Hu, Mater. Today Chem., 2025, 47, 102808.
16. M. Massetti, F. Jiao, A. J. Ferguson, D. Zhao, K. Wijeratne, A. Würger, J. L. Blackburn, X. Crispin and S. Fabiano, Chem. Rev., 2021, 121, 12465–12547.
17. D. Zhao, A. Würger and X. Crispin, J. Energy Chem., 2021, 61, 88–103.
18. M. Yu, H. Li, Y. Li, S. Wang, Q. Li, Y. Wang, B. Li, K. Zhu and W. Liu, EnergyChem, 2024, 6, 100123.
19. Y. Liu, M. Cui, W. Ling, L. Cheng, H. Lei, W. Li and Y. Huang, Energy Environ. Sci., 2022, 15, 3670–3687.
20. L. Yang, Y. Zhou, J. Xu, X. Ma, J. Yuan and B. Yuan, Int. J. Biol. Macromol., 2024, 282, 136881.
21. S. Jia, W. Qian, P. Yu, K. Li, M. Li, J. Lan, Y.-H. Lin and X. Yang, Mater. Today Phys., 2024, 42, 101375.
22. S. Sun, M. Li, X.-L. Shi and Z.-G. Chen, Adv. Energy Mater., 2023, 13, 2203692.
23. A. A. Quintana, A. M. Sztapka, V. de C. Santos Ebinuma and C. Agatemor, Angew. Chem., Int. Ed., 2022, 61, e202205609.
24. Y. Zhou, L. Yang, J. Xu, Z. Wei, X. Ma and B. Yuan, Int. J. Biol. Macromol., 2025, 30, 140324.
25. S. Kim, M. Ham, J. Lee, J. Kim, H. Lee and T. Park, Adv. Funct. Mater., 2023, 33, 2305499.
26. M. Fu, Z. Sun, X. Liu, Z. Huang, G. Luan, Y. Chen, J. Peng and K. Yue, Adv. Funct. Mater., 2023, 33, 2306086.
27. Y. Lei, Q. N. Chan, L. Xu, E. W. M. Lee, Y. X. Lee, V. Agarwal, G. H. Yeoh and W. Wang, Adv. Compos. Hybrid Mater., 2025, 8, 112.
28. Y. Chen, G. Hong, L. Li, Q. Qu, G. Li, J. Wu and L. Ge, Chem. Eng. J., 2024, 483, 149344.
29. D. Song, C. Zhao, B. Chen, W. Ma, K. Wang and X. Zhang, Joule, 2024, 8, 3217–3232.
30. S. Jia, T. Xiang, Z. Zhang, J. Zhou and L. Li, Chem. Eng. J., 2025, 503, 158563.
31. C. Lee, S. Hong and C. Liu, Macromol. Rapid Commun., 2025, 46, 2400837.
32. Z. Zhang, Z. Zhou, J. Huang and Y. Wang, J. Mater. Chem. A, 2024, 12, 6050–6066.
33. C. Jiang, X. Lai, Z. Wu, H. Li, X. Zeng, Y. Zhao, Q. Zeng, J. Gao and Y. Zhu, J. Mater. Chem. A, 2022, 10, 21368–21378.
34. Q. Cen, S. Chen, D. Yang, D. Zheng and X. Qiu, ACS Sustain. Chem. Eng., 2023, 11, 4473–4484.
35. M. Zhang, Q. Fu and H. Deng, Chem. Eng. J., 2024, 486, 150307.
36. H. Jia, Z. Ju, X. Tao, X. Yao and Y. Wang, Langmuir, 2017, 33, 7600–7605.
37. Q. Liu, X. Xu, Y. Zhang, L. Liang, B. Zhang and S. Chen, Chem. Eng. J., 2025, 509, 161207.
38. A. Lis, K. Zazakowny, K. Wolski, S. Zapotoczny and K. T. Wojciechowski, Chem. Eng. J., 2025, 515, 163468.
39. Y. He, S. Li, R. Chen, X. Liu, G. O. Odunmbaku, W. Fang, X. Lin, Z. Ou, Q. Gou, J. Wang, N. A. N. Ouedraogo, J. Li, M. Li, C. Li, Y. Zheng, S. Chen, Y. Zhou and K. Sun, Nano-micro Lett., 2023, 15, 101.
40. Z. Fan and J. Ouyang, Adv. Electron. Mater., 2019, 5, 1800769.
41. X. Yang, Y. Tian, B. Wu, W. Jia, C. Hou, Q. Zhang, Y. Li and H. Wang, Energy Environ. Mater., 2022, 5, 954–961.
42. A. Sohrabi, S. Liu, E. Cuce, Y. Shen, S. Y. Khan and M. Kumar, Renewable Sustainable Energy Rev., 2025, 219, 115832.
43. Z. Yu, X. Qu, Y. Wan, Q. Jiang, Y. Qin, J. Xu, J. Liu and H. He, Chem. Eng. J., 2024, 496, 154033.
44. L. Xue, H. Pan, Y. Wang, Z. Wang, X. Wang, Q. Lin, K. Zhang, X. Bu, M. Bai and B. Hong, Chem. Eng. J., 2025, 504, 158828.
45. Y. Zhao, Q. Zeng, C. Jiang, X. Lai, H. Li, Z. Wu and X. Zeng, Nano Energy, 2023, 116, 108785.
46. W. Zhao, C. Gao, F. Guo and Y. Wu, Res. Chem. Intermed., 2016, 42, 2953–2961.
47. H. Zuo, M. Wen, J. Shi, H. Park, L. Lv, Y. Ren, X. Zhao, H. Du, X. Yang and Z. Jian, Constr. Build. Mater., 2025, 490, 142465.
48. Y. Kong, H. Jin, G. Zhang and B. Yuan, Chem. Eng. J., 2024, 496, 154158.
49. Z. Xiong, W. Wang, Z. Wu, Z. Zhao, Q. Zeng, H. Li, X. Zeng, B. Wu and X. Lai, J. Mater. Chem. A, 2025 10.1039/D5TA04718H.
50. H. Jia, Y. Guo, Z. Chen, J. Wang, M. Döring, S. Liang and P. Jiang, Polym. Degrad. Stab., 2025, 239, 111393.

---

Footnote

† These authors contributed equally to this work and should be considered co-first authors.

---