Ferroelectric nanofibers with nanoconfined tellurium nanobeams for mechanical and thermal energy harvesting and wearable healthcare

Abstract

Large-area sensing devices open up great potential to trigger continuous self-powering, multifunctional sensors with responses to multiple stimuli. Herein, we report a self-powered flexible hybrid piezo- and pyro-electric nanogenerator (NG) based on Te-reinforced poly(vinylidene fluoride) (PVDF) electrospun nanofibers. It is tested to generate electricity from waste mechanical and thermal energies at room temperature. An electroactive phase (∼94%) is built in along the interface (an electrified jet) of dipoles poled perpendicular to the PVDF backbone ‘–C–C–’ chains. As a proof-of-concept, the fabricated NG generates remarkable electricity, reaching a power density of 4.2 μW cm⁻² under periodic mechanical stimulation, showcasing its potential for the efficient conversion of ambient mechanical energy into electricity. The intrinsic photothermal heat-localization effect of poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS) as an organic electrode enhances the pyroelectric response, yielding a pyroelectric coefficient of 40 μC m⁻² K⁻¹ (which is 4 times higher than that of the undoped fibers), making it capable of detecting low thermal oscillations. Therefore, it has the potential to monitor physiological conditions and body temperature, making it suitable for remote infectious disease surveillance.

---

New concepts

We introduce a self-powered, flexible, pressure- and temperature-responsive hybrid sensor based on tellurium (Te)-reinforced poly(vinylidene fluoride) (PVDF) electrospun nanofibers. This design strategically harnesses Te, a monoelemental chalcogenide with exceptional piezo- and ferro-electricity, to precisely modulate exotic charges within a 1D confinement. This approach not only boosts energy harvesting capabilities but also capitalizes on the synergistic effects of mechanical stretching and nanoconfined stress, significantly enhancing the alignment of molecular dipoles and optimizing ferroelectric properties, which are of prime importance in the era of sensor-based technologies. Our work demonstrates aligning the self-confined hexagonal phase of monoelemental Te along the fiber axis by utilizing high rotational speeds during electrospinning, thereby inducing a compressive lattice strain at the Te-PVDF interface. The quantum confinement of the Te nanobeams subtly alters their vibrational modes, fine-tuning the electronic states and significantly enhancing the charge-carrier coupling with the PVDF dipoles within the hybrid fibers. This process culminates in a robust heterojunction that amplifies both the piezoelectric and pyroelectric responses. Furthermore, we leverage organic PEDOT:PSS/Xyl electrodes to intensify photothermal heat localization compared to conventional metal electrodes, which markedly enhanced the pyroelectric response. This design achieves superior flexibility and durability while enabling highly sensitive pressure and temperature detection, establishing a new benchmark for self-powered wearable healthcare and remote infectious disease monitoring.

Introduction

Self-powered pressure sensors and electronic skin (e-skin) based devices have garnered growing attention recently in healthcare monitoring, biometrics and the ‘Internet of Things’ (IoT).^1–13 The idea of ‘internet connectivity’ is expanding beyond the traditional computer platforms of physical equipment and routine medical diagnosis tools.^6,10 Efforts are being directed to the development of smart ‘energy conversion’ devices based on self-sustaining piezo-, pyro- and ferro-electrics integrated with a hybrid nanogenerator to efficiently and simultaneously convert mechanical and thermal energies into a powerful electrical signal at room temperature. Several compatible dopants have been used to promote the functionalities of wearable devices.^1,14 Fragility limits the utility of inorganic doping in such devices. Thus, despite their inferior piezo- and pyro-electric features, organic polymers are being used in smart sensors and NGs owing to their good compatibility, good thermal stability, flexibility, inertness in ambient air, low cost, and overall good eco-friendliness and biocompatibility.^6,15

PVDF, a linear chain polymer and its light-weight copolymers are highly flexible, thermoplastic and electroactive with good piezo-, pyro- and ferro-electric properties that make them useful for smart electronic devices.^10–12 Crystalline PVDF exists in three major configurations: a non-polar α-phase, a polar β-phase, and a semi-polar γ-phase based on the polarities of ˙CH₂ and ˙CF₂ on its ‘–C–C–’ backbone chains.^10,16 The β-PVDF phase is the most electroactive ferroelectric (FE) phase for incorporation in piezo- and pyro-electric NGs, with the largest spontaneous polarization (P_s) and piezo-/pyro-electric coupling coefficients.^1,17 Mechanical stretching,^17,18 high electric field poling,^17,19 spin coating,²⁰ and electrospinning^1,17 are used to stretch and induce the electroactive β-PVDF phase. Among them, electrospinning is highly efficient to stretch the 1D β-PVDF configuration (D: dimension) with the dipoles aligned on the PVDF macromolecular chains, as the stretching force coupled with the electric field induces in situ polarization of duly enhanced piezoelectric properties. β-PVDF chain stretching under uniaxial stress self-aligns the dipoles, poled transverse to the stretching, resulting in enhanced piezoelectric potential and thereby efficiency. Nonetheless, due to the modest energy difference, the β phase relaxes back to α-PVDF.¹⁷ Heterojunctions, built-in with a strong electric field via fillers—such as graphene, graphene oxides (GO),^12,17 metals,²¹ carbon nanotubes (CNTs),^12,22 metal sulfides/oxides,^23–25 CsPbI₃ type halides,²⁶ Ag₂CO₃,¹⁴ and 2D transition metal dichalcogenides (TMDCs)²⁴—have shown to keep the dipoles mutually aligned perpendicular to the β-PVDF chains, which also restrains their relaxation back to the non-polar α-PVDF phase. In particular, collective electrostatic interactions between localized charges (fillers) and poled dipoles (polymer) at the heterojunction promote the stabilization of the β-PVDF phase hybrid nanostructure.

Molecular engineering of 2D architectures beyond GOs and TMDCs is a simple strategy for tuning exotic charges confined in a 2D-limited space.^27,28 For example, single layers of group III to group VI elements have led to breakthrough electronic materials with game-changing properties. In particular, the chalcogen Te possesses excellent piezo-, ferro- and pyro-electricity,^29–31 good thermoelectric^32,33 and non-linear optical behaviors,³⁴ and gate-tunable spin–orbit interactions useful for neuromorphic computing,²⁸ photodetectors,^35–37 and energy harvesting devices.^27,28 One challenge is the fabrication of large-area sensors for healthcare, biometrics, and other applications. In this scenario, Te nanocrystals (NCs) have multiple effects on the nucleation of the electroactive phase and energy harvesting functionalities in hybrid PVDF composites, such as: (i) stretching the PVDF polymer chains, (ii) poling the dipoles at the chains, (iii) increasing exotic surface charges, and (iv) coupling the charges (also spins) and dipoles regulated at cohesive nearest-neighbour interactions in a self-powered device. The two phases create a built-in robust structure (elastomer-like) in which the stretching force acts along the electrified jets. Thus, both piezo- and pyro-electric phases can be integrated into a single NG device. The open issues are to develop the science and technologies of self-powered inorganic–organic hybrid devices.

Herein, we report electrospun nanofibers of Te-reinforced PVDF polymer, which binds the Te NCs to be self-aligned along the fibers. A self-powered NG (the first report of its type) is fabricated from the fibers with built-in ferro-, piezo- and pyro-electric responsiveness for tracking physiological diseases and human body temperature in the context of remote infectious disease monitoring applications. The electroactive β/γ-PVDF phase is enhanced as the ˙CH₂ and ˙CF₂ dipoles are poled at the chains aligned along the fibers. The cohesive Te-PVDF interactions pave the dipole orders at the interfaces. A device with compatible organic electrodes and a pyroelectric coefficient enhanced by four orders of magnitude (p of ∼40 μC m⁻² K⁻¹) precisely detects high-temperature fluctuations. The device is tested for a wide range of pressure–temperature sensing with good flexibility, durability, cyclic stability, and electrical output, proving its potential in detecting physiological responses and variations in temperature, expanding its applications in real-time health care monitoring.

Results and discussion

Fig. 1a presents the X-ray diffraction (XRD) pattern of the single-phase small crystallites of α-Te (average size D_c = 30 nm) with a hexagonal crystal structure,^38,39 which were used to dope the PVDF polymer to form the hybrid nanocomposite of electrospun fibers. Small α-Te crystallites are critical to uniformly incorporate into PVDF fibers and regulate the Te-PVDF interactions over sufficiently large interfaces. The interplanar d_hkl spacing in the XRD peaks indicated average lattice parameters of a = 0.4443 nm and c = 0.5913 nm, with a lattice volume V_c = ½(3)^1/2a²c ≅ 0.1011 nm³ and X-ray crystal density ρ_c = 6.286 g cm⁻³. The small α-Te crystallites are well surface-compressed (strained) upto 0.90% larger density compared to the bulk density of 6.230 g cm⁻³. The α-Te lattice has been shown to contract in 1D nanowires.⁴⁰ Further, the present α-Te exhibits anisotropic growth, demonstrated by its enhanced (100) peak intensity (relative to the most intense (101) peak) of I₁₀₀/I₁₀₁ = 28%, compared to a value of I₁₀₀/I₁₀₁ = 16% in the bulk state. Primarily, lateral growth dominates along the a-axis. The I₁₀₀ value is further enhanced to 87% in the Te-reinforced PVDF fibers, which shows that the α-Te is aligned along the fibers (Fig. S1). A compressive strain γ_c ≅ (2d₂₀₀ − d₁₀₀)/d₁₀₀ = (−) 2.29% was estimated (shown in the inset) from the interplanar spacings of d₁₀₀ = 0.3448 nm and d₂₀₀ = 0.1888 nm observed for the respective peaks. The field-emission scanning electron microscopy (FE-SEM) images in Fig. 1b display long nanoprisms (or nanobeams) (w ≅ 200–500 nm widths) grown along the (001) facets of the model given at the right. The nanobeams contain large surfaces that form effectively large interfaces when doped in the PVDF fibers.

Fig. 1 (a) XRD pattern of α-Te nanobeams, and their (b) FE-SEM images (models at right) and (c) Raman bands with schematic energy levels shown in the inset. (d) The three A₁ bands split up in Te atoms, oscillate along and perpendicular to the nanobeams. (e) A TEM image of a PVTe_1.0 fiber with embedded α-Te and (f) d₁₀₀ lattice images of α-Te bound by the d_110* polymer chains via an interface, as shown in (g) a model structure.

In the Raman spectrum in Fig. 1c (full spectra are provided in Fig. S2), the Raman bands of three well-known groups of E₁ (Te–Te bending), A₁ (Te–Te breathing) and E₂ (asymmetric Te–Te stretching along the c-axis) modes of Te atoms oscillating in a chain are observed at 103, 125 and 147 cm⁻¹. Duly enhanced phonon frequencies, if compared to thin Te@SiO₂ films (94, 122, and 142 cm⁻¹, respectively),³⁵ anticipate the α-Te is quantum confined, as also evidenced from the XRD pattern. As illustrated in Fig. 1d, the three A₁ bands v₁, v₂ and v₃ are located at 116, 125 and 131 cm⁻¹ and correspond to a transverse, a longitudinal, and a mixed oscillation, respectively, of the Te atoms in the nanobeams. Additionally, the E₂ band exhibits three parts (marked in a schematic energy level diagram in the inset in Fig. 1c). No asymmetry is visible in the E₁ mode. Coupled phonons with the valence 5s²p⁴ Te electrons (in the Te-atoms oscillating at the triplet ground electronic state ³S₁) results in the multiple states.

The typical transmission electron microscopy (TEM) image in Fig. 1e shows how a reinforced fiber PVTe_1.0 embeds α-Te of small rhomboids, which are grown edge-on towards the (001) facets lying along the PVDF fibers. This means the PVDF dictates the α-Te to order, in support, on and along a stable interface of the Te atoms bonded to the ˙CH₂ and ˙CF₂ dipoles along the fibers. A higher-magnified view reveals that the d₁₀₀ = 0.3450 nm Te lattices (Fig. 1f) are inclined at ϑ = 32°, bonding to the PVDF β-phase at the d_110* = 0.3825 nm chains via an interface (disordered, w = 2.0 nm) in an integrated structure. Fig. 1g illustrates a model in which the two phases are mutually ordered via an interface in a fiber. The aligned d_110* polymer chains are reflected in the duly enhanced intensity of the d_110* XRD peak (Fig. S1b). The FESEM images in Fig. 2a and b present bead-free fibers (insets show histograms of the fibers with uniformly distributed diameters) electrospun at R_s = 500 rpm (w = 246 ± 13 nm) and 3000 rpm (w = 235.4 ± 17 nm), respectively. Relatively narrow distributions are observed for the fibers processed under these two conditions, which confirms their homogeneity across the macroscopic area of the sample. A large fraction φ → 90%, of oriented fibers is achieved at the optimized R_s of 3000 rpm (≤15% at R_s = 500 rpm) (Fig. 2c and d). The fast Fourier transform (FFT) images (insets) present diverged intensities at all angles (a Gaussian-shaped curve) in the randomly oriented fibers, while majority of the fibers are oriented at 92° when electrospun at the faster R_s. The pixels, scattered in an arc in the randomly oriented fibers, converged in an ellipsoid in the preferentially aligned fibers. Elemental mapping (Fig. 2e) confirmed that C, F and Te were uniformly distributed in the α-Te reinforced PVTe_1.0 fibers. In the PVTe nanofibers (Fig. 2f, g and h), the (001) planes of the α-Te NCs face up along thick fibers, (w ≥ 150 nm), while the lateral α-Te faces are set along thinner fibers, as projected in the models.

Fig. 2 FESEM images of the PVTe_1.0 fibers (distributions of their diameters are shown in histograms in the insets) electrospun at (a) 500 rpm and (b) 3000 rpm and (c and d) their different fractional (φ) orientations, respectively, with the FFT images in the insets. (e) A typical TEM image of a Te-nanobeam (rhomboid) embedded in a PVDF fiber, with the C, F and Te mapping shown at right. A closer view of (f) randomly oriented and (g and h) aligned fibers, showing that the Te NCs are embedded either along the (001) facets, or the lateral facets.

The FTIR bands (Fig. 3a, full spectra are provided in Figures S3 and S4) reveal successive increase in the β-PVDF phase fraction with α-Te doping (plotted in Fig. 3b). A quantitative assessment of the concentration dependence of the β-PVDF phase fraction on the Te doping (Fig. S5, associated discussion S2) shows a maximum F_β of 82% (electroactive fraction F_EA ∼94% at R_s = 3000 rpm). The characteristic β-PVDF phase bands at 840 and 1277 cm⁻¹ are enhanced in PVTe_1.0 at the expense of the α-PVDF phase bands at 763, 796 and 977 cm⁻¹, which were used to estimate F_EA and F_β as follows:⁴¹

where A_EA840 and A_α763 are the absorbance at the 840 cm⁻¹ and 763 cm⁻¹ bands with the absorption coefficients K_EA840 and K_α763 (K_EA840/K_α763 ∼ 1.26), respectively, and A₁₂₇₇ and A₁₂₃₄ are the absorbance intensities at 1277 and 1234 cm⁻¹, which correspond to the β- and γ-PVDF phases, respectively. Thus, the maximum F_β and F_EA (Table S1) are tuned in uniaxial stretching along the applied field E and stress, with the dipoles poled perpendicular to the polymer chains.^42,43 A cohesive interaction of surface Te charges with β-PVDF dipoles results in softer ˙CH₂ stretching bands (Fig. S6) due to damped oscillation with the enhanced effective ˙CH₂ mass,¹⁴where 2r_dc is the damping constant, f₀ and f_int are the frequencies (˙CH₂ stretching) before and after the doping, respectively, and c is the velocity of light. A 2r_dc value of ∼5.8 × 10¹³ s⁻¹ was thus estimated for the Te-doped fibers (Fig. 3b).

Fig. 3 (a) FTIR spectra (* signifies bands common to the different phases) of electrospun NPV and PVTe fibers (R_s = 3000 rpm), and (b) variation in fractions (%) F_α/F_β of the α-/β-phases and damping factor 2r_dc with the Te content. (c) Schematic representations of the polymer chain axis (–C–C– bond) and transitional dipole moments (μ) along the a, b and c-axes with respect to the direction of the polarizer. (d) FTIR spectra of the aligned nanofibers measured using parallel and perpendicular polarized IR beams. (e) C 1s, (f) F 1s and (g) Te 3d XPS bands of the fibers.

The alignment of the dipoles in the fibers are observed as polarized FTIR vibrational bands. As shown in the schematics in Fig. 3c, if the fibers are stretched towards the c-axis along the rotation of the collector (Fig. S7), the polymer chains would be oriented along the rotation that makes an angle at the c-axis. If the c-axis lies parallel to the polarized IR beam, the component of the transition dipole moment (μ_c) along the c-axis induces intensified bands (Fig. 3d, full spectra provided in Fig. S8) of ω(CH₂) + ν_as(CC) at 1404 cm⁻¹, and ν_s(CC) + ω(CF₂) at 1070 cm⁻¹ in a parallel mode (ω = wagging and ν_as (s) = asymmetric (symmetric) stretching). In a perpendicular polarized IR beam, the component μ_b along the b-axis induces intensified bands of ν_s(CF₂) + ν_s(CC) at 1277 cm⁻¹ and ν_s(CF₂) + ν_s(CC) at 840 cm⁻¹, with decreased intensities of the 1404 cm⁻¹ and 1070 cm⁻¹ bands (as listed in Table S2).⁴⁴ This indicates the preferential molecular chain orientation and improved dipole alignment at faster R_s compared to slower R_s (∼83% and ∼54%, respectively, in PVTe_1.0) (Figures S9, S10, associated discussion S3). The PVDF β-phase has a 2.3% shorter d₁₁₀ = 0.4285 nm (PVT_1.0) in the XRD pattern (Fig. S1) compared to the virgin fibers (d₁₁₀ = 0.4385 nm). The XPS C 1s bands of the ˙CH₂ and ˙CF₂ moieties (Fig. 3e) show higher binding energy (BE) by ΔE_c = 0.35 eV and 0.20 eV, respectively, over the virgin states, with a CH₂ band intensity enhancement Δh_CH2 of ∼ 5%. The reduced gap of 4.45 eV (4.60 eV before doping) between the two C 1s bands indicates Te-mediated Te-PVDF charge–dipole interactions. In Fig. 3f and g, the F1s band is also blue shifted by 0.1 eV, and the interface Te-3d_3/2,_5/2 bands have values 2.15 eV higher than those of the primary bands. Accordingly, the Fermi surface is pushed closer to the conduction band in a finely tuned band structure. Bare Te (Fig. S11) has 0.8 eV lower Te-3d_3/2,_5/2 values in the absence of interactions with ˙CH₂ and ˙CF₂ dipoles at the polymer chains.

At frequencies below 10⁴ Hz, PVTe_1.0 (Fig. 4a) retains a steady dielectric permittivity ε_r of ∼16 (11 before doping) at a well-controlled power loss (tan δ) of ≤0.02, making it usable for applications over wide bands of frequencies. A reasonably enhanced ε persists at the dipoles aligned at the Te-PVDF interfaces due to the Maxwell–Wagner–Sillars (MWS) effect.⁴⁵ A stable interface regulates the charge order of the enhanced ε_r values that govern the piezo- and ferro-electric features.²⁶ A peak is observed near tanδ = 2 × 10⁶ Hz due to the rather metallized states induced on the charge-order. Fig. 4b and c display the switching response of PVTe_1.0 (deposited on ITO glass) to an AC voltage applied between the probe and the substrate with a ±10 V DC bias. At positive bias, the phase-hysteresis shows that the dipoles are switched within the fibers up to a (−) 90° phase reversal to attain saturation. Under reversed fields, they reorient oppositely up to a 90° phase value in the local FE response. The enhanced magnitude of piezoelectric coefficient (13±2 pm V⁻¹ before doping, Fig. S12), at a change of Δz in the lattice strain (amplitude response) with a change of ΔV in the applied field,^8,46 indicates that they could be useful for developing self-powered devices. This value is superior (see comparison in Table S3) to those of most advanced piezoelectric materials.

Fig. 4 (a) Permittivity (ε_r) and power loss (tan δ) for the NPV and PVTe_1.0 nanofibers. (b) Ferroelectric hysteresis loop (PVTe_1.0) showing a 180° phase reversal at switching of the dipoles, and (c) a piezoelectric butterfly hysteresis loop with d₃₃ = 21 ± 1 pm V⁻¹. (d) Schematic of the fabricated NG device used to generate a (e) V_OC at varied pressures, (f) I_sc profile at 128 kPa, (g) its electrical output voltage (V_L) and power density (P_d) at varied loads (equivalent circuit in the inset), (h) its mechano-sensitivity under different pressures, and (i) its long-term mechanical stability over 4500 cycles, which was maintained even after 6 months.

The PVTe_1.0 NG device depicted in the schematic in Fig. 4d was tested to validate its viability for energy harvesting applications. When subjected to an external mechanical stress, a piezo-potential develops across the device. Thus, a peak-to-peak open-circuit voltage V_oc of ∼8.5 V (≤2.4 V before doping, Fig. S13a) was obtained (Fig. 4e) using an imparting machine (with a motor used as the source of impact) under a periodic applied pressure (at a frequency of 10 Hz) of up to 128 kPa, with a short-circuit current I_sc of ∼1.2 μA (Fig. 4f) (∼0.53 μA before doping, Fig. S13b). Furthermore, the device demonstrates a rapid response time of ∼12 ms, as determined from the output voltage profiles under different applied pressure levels (Fig. S14). A common switching-polarity test validated the piezoelectric nature of the sign of V_oc, which reverses under forward and reverse conditions under an equivalent imparting pressure of 68 kPa (Fig. S15). The instantaneous voltage V_L gradually increased to 5.1 V (Fig. 4g), giving rise to a maximum power density P_d of ∼4.2 μW cm⁻² at the critical load resistance R_L = 0.8 MΩ across the device with an effective area of A,

These values represent marked improvements over previously reported values (Table S4). The mechano-sensitivity S = ∂V/∂σ_a shows two linear pressure response regimes (Fig. 4h): the former with mechano-sensitivity of 101 mV kPa⁻¹ below 30 kPa, enabling the detection of pressures as low as 5 kPa; lowered to 21 mV kPa⁻¹ in the latter regime due to the theoretical limit of the effective strain in piezoelectric materials in the high-pressure region. The mechanical durability and stability prolong safely for up to 4500 cycles, with no significant change in the amplitude of the voltage signal under a constant imparting force throughout the extended testing period, even after 6 months (Fig. 4i). This validates the long-term operational stability of the device even in harsh conditions.

The sensing performance of the device was tested towards different mechanical deformation, human motion and vibration stimuli.^5,47 To demonstrate its versatility and tolerance, it was placed under mechanical deformation by bending at 30°, 60° and 90° angles. A quick bending motion (in an arc) generates V_oc, which increases with increasing bending towards a maximum of 90° (Fig. 5a). The sensing behaviour of the device was systematically evaluated under realistic physiological conditions. In due flexibility and sensitivity towards bending, it converted several physiological motions (illustrated in Fig. 5b) into a V_oc signal upon deformation. The pattern differs (Fig. 5c) for the bending of different synovial joints, i.e., the finger, neck, elbow and wrist. The dynamic parameters can be utilized in diagnosing various infectious diseases.⁷ Our device showed repeatable sensing of the thoracic pressure developed in the muscle tissues around the vocal cords during the action of coughing at the larynx (Fig. S16). The results demonstrate its stability and reproducibility towards subtle changes in instantaneous pressure, thereby allowing the detection of tactile stimuli and vocal cord vibrations for a wide range of real-time applications.

Fig. 5 (a) Electrical outputs of the flexible NG used for mechanical sensing when bent at different angles. (b) A schematic diagram illustrating pertinent physiological gesture detections during ‘bending’ of the finger, neck, elbow and wrist, and the corresponding (c) induced electrical output signals.

To investigate its pyroelectric performance, a device architecture was engineered using PEDOT:PSS/Xyl composite electrodes. This material system was selected for its synergistic properties of high electrical conductivity, intrinsic photothermal conversion abilities, and mechanical flexibility. The primary objective of this study was to elucidate the role of the electrode composition in modulating the pyroelectric response dynamics under photothermal excitation. The pyroelectric response of the device was tested via IR heating at (Fig. 6a and b) at a 0.1 Hz switching frequency, which induces an I_sc of ∼0.4 nA at a faster response time, τ of 0.12 s (Fig. 6c and d), reported so far.⁴⁸ The decreasing charge polarization with rising temperature (T) induces a positive current response.⁴⁹ The reversed I_sc during subsequent cooling anticipates enhanced charge interactions,

where h is the height and w is the width of the nanowires.⁵⁰ The enhanced p value of 40 μC m⁻² K⁻¹ (10 μC m⁻² K⁻¹ in the bare fibers, see Fig. S17), with a significant V_oc of 145 mV (Fig. S18) demonstrates the merits the of the device as a thermal energy harvester and breathing sensor.⁴⁸ The observed enhancement arises from both the device engineering and the intrinsic thermal properties of the Te-doped PVDF fibers, which effectively regulate the pyroelectric response. To further elucidate the thermal transport characteristics, the specific heat capacities (C_p) of both the undoped and doped fibers (Fig. S19) were measured to calculate the volume-specific heat C_E = ρ_cC_p (where the density ρ_c is ∼1.74 g cm⁻³).⁵¹ Assuming similar thermal diffusivity values (α) for both fibers, the thermal conductivity κ = αC_E for PVTe_1.0 was estimated to be significantly lower (∼0.55 W m⁻¹ K⁻¹) than that of NPV (0.91 W m⁻¹ K⁻¹).⁵²

Fig. 6 Pyroelectric performance of PVTe_1.0 using PEDOT:PSS/Xyl gel as the electrode. (a) Input temperature profile, (b) corresponding heat rate (at 0.1 Hz switching frequency), (c) induced I_sc during the heating and cooling cycle, and (d) a comparison of the pyroelectric current response-time of different pertinent materials, showing the fast response of our device. (e) Schematic diagrams of the working mechanism of the pyroelectric nanogenerator showing (i) the initial polarization state and change in the polarization state upon (ii) heating and (iii) cooling, causing alternate current flow in external circuit. (f) Comparison of the pyroelectric coefficients of different organic and inorganic materials (ref. 50 and 54–68).

The κ value is markedly reduced upon the doping of α-Te in the PVDF fibers, as it acts as a phonon-scattering centre, thus impeding heat transport.⁵³ The restricted heat flow promotes photothermal localization, resulting in the build-up of elevated local temperatures in the PVTe_1.0 nanofibers. Consequently, thermal gradients are retained more effectively, minimizing heat transfer to the surroundings. The localized thermal strain between the PEDOT:PSS/Xyl electrode and PVTe_1.0 interface is therefore more pronounced compared to that in the NPV system, leading to a notable boost in the pyroelectric output. Moreover, the PEDOT:PSS/Xyl film electrode demonstrates a notably higher absolute temperature (T) and an increased temperature difference (ΔT) of 12 K, along with a higher dT/dt magnitude compared to the metal electrode (Fig. S20) (for associated discussion, see S4). This observation confirms the superior photon energy absorption efficiency of our system with organic electrodes.

Pyroelectric materials demonstrate a temperature-dependent spontaneous polarization (P) arising from their non-centrosymmetric crystal structure, which generates surface charges in the absence of an external electric field. At thermal equilibrium (dT/dt = 0), the static polarization induces bound surface charges that are neutralized by adsorbed environmental ions or electrons, resulting in no net current flow through externally connected electrodes (Fig. 6e(i)). This equilibrium persists until thermal perturbations disrupt the system. When subjected to heating (dT/dt > 0), the increased lattice thermal energy disrupts the alignment of the intrinsic electric dipoles within the material. This disordering reduces the spontaneous polarization (ΔP/ΔT < 0), thereby decreasing the surface charge density. The resulting excess compensation charges are expelled from the surfaces of the material, generating a measurable pyroelectric current in the external circuit (Fig. 6e(ii)). Conversely, during cooling (dT/dt < 0), the lattice contracts and the dipoles align, leading to an increase in the P value. This reordering elevates the surface charge density as free charges migrate from the environment to re-establish the electrostatic equilibrium (Fig. 6e(iii)). The resultant reverse current, flowing under short-circuit conditions, confirms the reversibility of the pyroelectric effect, highlighting its dependence on the direction and magnitude of the thermal fluctuations. It is worth mentioning that our device exhibits a significantly higher p value than those that have been reported for pyroelectric devices based on organic and inorganic materials (Fig. 6f).^54–68

Conclusions

Te-reinforced PVDF nanofibers have been shown to exhibit enhanced pyro-electric features along with a piezo-electric response suitable for the fabrication of wearable pressure and temperature sensors. Optimized doping with 1 wt% Te (nanobeams) leads to poled, greatly promoted electroactive β/γ-phases at the interfaces, with a highly enhanced d₃₃ value of 21 ± 1 pm V⁻¹ at a 10 V bias. The device generates an 8.5 V signal (4.2 μW cm⁻² power density), sufficient to run self-powered electronics. As a proof-of-concept, it was tested in the detection of a variety of human physiological signals, including the motions of joints during bending and the changes in thoracic pressure during coughing, showing its potential in biomedical diagnosis and voice recognition systems. Its pyroelectric features surpass those of bare PVDF fibers to detect minor thermal fluctuations, giving it great potential for real-world application in multimodal smart sensors for acoustic and medical devices.

Experimental details

Synthesis

The α-Te NCs were prepared from Na₂TeO₃ using a reduction reaction in hydrazine (N₂H₄·H₂O) at 180 °C for 36 h (for associated discussion, see S1). After the reaction (2Na₂TeO₃ + N₂H₄·H₂O → 2Te + 4NaOH + H₂O + 2NO↑), the recovered Te sample was dispersed in a solution of PVDF in N,N-dimethylformamide (DMF) and acetone (6 : 4 v/v ratio) via stirring at 60 °C to obtain a homogeneous precursor of nanocolloids. Hydrazine is a mild reducing agent that yields tiny α-Te NCs dispersed in aqueous NaOH as a slurry.³⁵ The polar liquid DMF mediates the dispersion of the α-Te NCs on the PVDF chains in the nanocolloids. The sample was injected through a syringe under a voltage of 14 kV between a metallic needle tip and an aluminum-foil-wrapped drum collector (separation: 12 cm), at which it was electrospun as nanofibers (Fig. S7). Using this method, two samples of Te-reinforced PVDF containing 0.5 wt% (PVTe_0.5) and 1.0 wt% Te (PVTe_1.0) were obtained (at drum rotation speeds of R_s = 500 rpm and 3000 rpm) as hybrid structures with functional properties.

Characterizations

The electroactive β/γ-PVDF phases of the oriented dipoles formed in the nanofibers were assessed using Fourier transform infrared (FTIR) spectra recorded using a Thermo Scientific, iS20 Nicolet instrument with a KRS-5 polarizer. Their crystallographic structures were analyzed from the XRD patterns recorded using an X-ray diffractometer with CuKα X-ray radiation (λ = 0.15410 nm). Field emission scanning microscopy (FE-SEM) and high-resolution transmission electron microscopy (HRTEM) were used to study the surface topologies of the bare α-Te-NCs, PVDF fibres and Te-PVDF nanocomposite fibers. The XPS analysis was performed using a Thermo Scientific K-Alpha spectrophotometer equipped with a monochromatic Al Kα X-ray source with an energy of hν = 1486.67 eV. The system employs a dual-beam charge compensation system that uses both a low-energy electron flood gun and an Ar ion flood gun to prevent local electrostatic charge build-up on the sample surface. The binding energies of the XPS bands were corrected with respect to the adventitious carbon C 1s band at 284.8 eV. The dielectric properties were measured using an impedance analyzer (E4990A, Keysight). The PFM (piezoelectric force microscopy) signal was studied using a Nanosurf CoreAFM system, and the electrical response of the NG device was collected using a digital storage oscilloscope (Keysight, DSOX1102G) and electrometer unit (Keithley, 6514). Other details are the same as reported previously.⁴⁹

Fabrication

The NG for examining piezoelectric performance consisted of a nanofiber mat (2 × 5 cm²) sandwiched between the top and bottom electrodes of a synthetic Ni–Cu textile, which was then machined to remove air gaps so the electrodes would form contacts with the mat connected in a circuit. Finally, as reported earlier,⁴⁹ the device was covered in 400-μm-thick polydimethylsiloxane using the elastomer Sylgard 184 to protect it from mechanical damage. For the preparation of the PEDOT:PSS/Xyl electrodes, a suspension of PEDOT:PSS was combined with a few drops of 4N HCl, followed by the addition of xylitol pellets in a weight ratio of 6 : 1 (w/w). The resulting mixture was stirred for 1 h and then placed in a sealed water bath at 100 °C for 2 h. The PEDOT:PSS/Xyl film electrodes were then fabricated by spreading the synthesized gel onto both sides of a nanofiber mat of the desired size and heating the assembly at 80 °C in air.

Author contributions

US synthesized the material, executed the experiments for material and device characterization and wrote the MS. HKM planned and carried out the device measurements, provided critical feedback and revised the MS. AK carried out PFM measurement and contributed to data analysis. SR contributed to data analysis of data obtained from TEM and XRD and revised the MS. DM planned and supervised the entire work. The manuscript was written with the contributions of all the authors.

Conflicts of interest

There are no conflicts to declare.

Data availability

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

Supplementary information: The supporting information contains associated discussions and detailed synthetic procedures, XPS spectra, a table of the calculated electroactive phases, FTIR data of the various –CH₂ stretching, a table of the assignments of the vibrational bands, comparison of the degrees of the oriented dipoles, XRD patterns, the PFM response of NPV, piezoelectric and pyroelectric output responses of NPV, PVTe_1.0 and the polarity reversal test. See DOI: https://doi.org/10.1039/d5mh01052g.

Acknowledgements

This work is supported by the ANRF (SERB/CRG/2020/004306), the Government of India, for financial support. US and AK are thankful to the University Grant Commission (UGC) for providing the fellowships.

References

1. C. Chang, V. H. Tran, J. Wang, Y. K. Fuh and L. Lin, Nano Lett, 2010, 10, 726–731.
2. Y. Yang, W. Guo, K. C. Pradel, G. Zhu, Y. Zhou, Y. Zhang, Y. Hu, L. Lin and Z. L. Wang, Nano Lett, 2012, 12, 2833–2838.
3. M. Amjadi, K. U. Kyung, I. Park and M. Sitti, Adv. Funct. Mater., 2016, 26, 1678–1698.
4. D. Y. Park, D. J. Joe, D. H. Kim, H. Park, J. H. Han, C. K. Jeong, H. Park, J. G. Park, B. Joung and K. J. Lee, Adv. Mater., 2017, 29, 1702308.
5. M. Lou, I. Abdalla, M. Zhu, X. Wei, J. Yu, Z. Li and B. Ding, ACS Appl. Mater. Interfaces, 2020, 12, 19965–19973.
6. R. A. Surmenev, R. V. Chernozem, I. O. Pariy and M. A. Surmeneva, Nano Energy, 2021, 79, 105442.
7. X. Huang, Q. Qin, X. Wang, H. Xiang, J. Zheng, Y. Lu, C. Lv, K. Wu, L. Yan, N. Wang, C. Xia and Z. L. Wang, ACS Nano, 2021, 15, 19783–19792.
8. H. K. Mishra, V. Gupta, K. Roy, A. Babu, A. Kumar and D. Mandal, Nano Energy, 2022, 95, 107052.
9. C. Ye, S. Yang, J. Ren, S. Dong, L. Cao, Y. Pei and S. Ling, ACS Nano, 2022, 16, 4415–4425.
10. C. M. Costa, V. F. Cardoso, P. Martins, D. M. Correia, R. Gonçalves, P. Costa, V. Correia, C. Ribeiro, M. M. Fernandes, P. M. Martins and S. Lanceros-Méndez, Chem. Rev., 2023, 123, 11392–11487.
11. P. Chen, C. Cheng, X. Yang, T. T. Sha, X. Zou, F. Zhang, W. Jiang, Y. Xu, X. Cao, Y. M. You and Z. Luo, ACS Nano, 2023, 17, 25625–25637.
12. A. Huang, Y. Zhu, S. Peng, B. Tan and X. Peng, ACS Nano, 2024, 18, 691–702.
13. A. Kumar, R. Haldar, E. E. Siddharthan, R. Thapa, M. Shanmugam and D. Mandal, Nano Lett., 2024, 24, 7861–7867.
14. H. K. Mishra, D. Sengupta, A. Babu, B. M. Pirzada, R. Sarkar, B. S. Naidu, T. K. Kundu and D. Mandal, Sustainable Energy Fuels, 2022, 6, 1625–1640.
15. S. Mishra, L. Unnikrishnan, S. K. Nayak and S. Mohanty, Macromol. Mater. Eng., 2019, 304, 1–25.
16. A. D. Phule, S. Ram, S. K. Shinde, J. H. Choi and A. K. Tyagi, J. Phys. Chem. Solids, 2018, 115, 254–264.
17. X. Liu, J. Ma, X. Wu, L. Lin and X. Wang, ACS Nano, 2017, 11, 1901–1910.
18. L. Li, M. Zhang, M. Rong and W. Ruan, RSC Adv, 2014, 4, 3938–3943.
19. A. Morali, A. Mandal, M. Skorobogatiy and S. Bodkhe, RSC Adv, 2023, 13, 31234–31242.
20. M. Fortunato, D. Cavallini, G. De Bellis, F. Marra, A. Tamburrano, F. Sarto and M. S. Sarto, Polymers, 2019, 11, 1096–1110.
21. A. Gebrekrstos, T. S. Muzata and S. S. Ray, ACS Appl. Nano Mater, 2022, 5, 7632–7651.
22. L. Yang, Q. Zhao, K. Chen, Y. Ma, Y. Wu, H. Ji and J. Qiu, ACS Appl. Mater. Interfaces, 2020, 12, 11045–11054.
23. T. Yang, H. Pan, G. Tian, B. Zhang, D. Xiong, Y. Gao, C. Yan, X. Chu, N. Chen, S. Zhong, L. Zhang, W. Deng and W. Yang, Nano Energy, 2020, 72, 104706.
24. D. Bhattacharya, S. Bayan, R. K. Mitra and S. K. Ray, Nanoscale, 2021, 13, 15819–15829.
25. X. García-Casas, F. J. Aparicio, J. Budagosky, A. Ghaffarinejad, N. Orozco-Corrales, K. (Ken) Ostrikov, J. R. Sánchez-Valencia, Á. Barranco and A. Borrás, Nano Energy, 2023, 114, 108686.
26. Y. Xue, T. Yang, Y. Zheng, K. Wang, E. Wang, H. Wang, L. Zhu, Z. Du, H. Wang, K. C. Chou and X. Hou, Adv. Sci., 2023, 10, 2300650.
27. N. R. Glavin, R. Rao, V. Varshney, E. Bianco, A. Apte, A. Roy, E. Ringe and P. M. Ajayan, Adv. Mater., 2020, 32, 1–22.
28. G. Rao, H. Fang, T. Zhou, C. Zhao, N. Shang, J. Huang, Y. Liu, X. Du, P. Li, X. Jian, L. Ma, J. Wang, K. Liu, J. Wu, X. Wang and J. Xiong, Adv. Mater., 2022, 34, 1–10.
29. Q. Lian, W. Liu, D. Ma, Z. Liang, Z. Tang, J. Cao, C. He and D. Xia, ACS Nano, 2023, 17, 8755–8766.
30. J. Zhang, J. Zhang, Y. Qi, S. Gong, H. Xu, Z. Liu, R. Zhang, M. A. Sadi, D. Sychev, R. Zhao, H. Yang, Z. Wu, D. Cui, L. Wang, C. Ma, X. Wu, J. Gao, Y. P. Chen, X. Wang and Y. Jiang, Nat. Commun., 2024, 15, 7648.
31. H. K. Mishra, A. Jain, D. Saini, B. Mondal, C. Bera, S. Ram and D. Mandal, Phys. Rev. B, 2025, 111, 155436.
32. A. Chauhan, A. K. Das, D. K. Satapathy and M. Battabyal, ACS Appl. Energy Mater., 2025, 8, 6044–6055.
33. D. Liu, D. He and H. Huang, Appl. Phys. Lett, 2025, 126, 043505.
34. J. Zha, D. Dong, H. Huang, Y. Xia, J. Tong, H. Liu, H. P. Chan, J. C. Ho, C. Zhao, Y. Chai and C. Tan, Adv. Mater., 2024, 36, 2408969.
35. Y. Du, G. Qiu, Y. Wang, M. Si, X. Xu, W. Wu and P. D. Ye, Nano Lett, 2017, 17, 3965–3973.
36. M. Amani, C. Tan, G. Zhang, C. Zhao, J. Bullock, X. Song, H. Kim, V. R. Shrestha, Y. Gao, K. B. Crozier, M. Scott and A. Javey, ACS Nano, 2018, 12, 7253–7263.
37. C. Niu, G. Qiu, Y. Wang, Z. Zhang, M. Si, W. Wu and P. D. Ye, Phys. Rev. B, 2020, 101(205414), 6.
38. Joint Committee on Powder Diffraction Standards, X-ray Powder Diffraction File No. 00-036-1452, α-Te (International Centre for Diffraction Data, Swarthmore, PA, 2019).
39. H. K. Mishra, A. Jain, D. Saini, B. Mondal, C. Bera, S. Ram and D. Mandal, Phys. Rev. B, 2025, 111(7), 155436.
40. H. S. Qian, S. H. Yu, J. Y. Gong, L. B. Luo and L. F. Fei, Langmuir, 2006, 22, 3830–3835.
41. J. Gomes, J. S. Nunes, V. Sencadas and S. Lanceros-Mendez, Smart Mater. Struct., 2010, 19, 065010.
42. M. M. Abolhasania, S. Azimia and H. Fashandi, RSC Adv, 2015, 5, 61277.
43. R. K. Singh, S. W. Lye and J. Miao, Polymer, 2021, 214, 123366.
44. D. Mandal, S. Yoon and K. J. Kim, Macromol. Rapid Commun., 2011, 32, 831–837.
45. A. Kumar and D. Mandal, J. Appl. Polym. Sci., 2023, 140, e53652.
46. A. Jalalian, A. M. Grishin, X. L. Wang, Z. X. Cheng and S. X. Dou, Appl. Phys. Lett, 2014, 104, 103112.
47. D. W. Lee, D. G. Jeong, J. H. Kim, H. S. Kim, G. Murillo, G. H. Lee, H. C. Song and J. H. Jung, Nano Energy, 2020, 76, 105066.
48. K. Roy, S. K. Ghosh, A. Sultana, S. Garain, M. Xie, C. R. Bowen, K. Henkel, D. Schmeiβer and D. Mandal, ACS Appl. Nano Mater, 2019, 2, 2013–2025.
49. D. Saini, D. Sengupta, B. Mondal, H. K. Mishra, R. Ghosh, P. N. Vishwakarma, S. Ram and D. Mandal, ACS Nano, 2024, 18, 11964–11977.
50. C. R. Bowen, J. Taylor, E. Leboulbar, D. Zabek, A. Chauhan and R. Vaish, Energy Environ. Sci., 2014, 7, 3836–3856.
51. Y. Zhang, M. Xie, J. Roscow, Y. Bao, K. Zhou, D. Zhang and C. R. Bowen, J. Mater. Chem. A, 2017, 5, 6569–6580.
52. Y. Zhang, X. Zhang, L. Yang, Q. Zhang, M. L. Fitzgerald, A. Ueda, Y. Chen, R. Mu, D. Li and L. M. Bellan, Soft Matter, 2018, 14, 9534–9541.
53. X. Cui and Y. Cai, ACS Appl. Nano Mater, 2023, 6, 17457–17463.
54. D. Berlincourt, H. Jaffe and L. R. Shiozawa, Phys. Rev., 1963, 129, 1009–1017.
55. M. H. Litt, C. H. Hsu and P. Basu, J. Appl. Phys, 1977, 48, 2208–2212.
56. V. S. Likhovidov, V. V. Golovanov and A. V. Vannikov, Polym. Sci. U. S. S. R., 1978, 20, 80–87.
57. S. B. Lang, A. S. DeReggi, M. G. Broadhurst and G. T. Davis, Ferroelectrics, 1981, 33, 119–125.
58. M. H. Lee, R. Guo and A. S. Bhalla, J. Electroceram., 1998, 2, 229–242.
59. V. Bharti and R. Nath, J. Phys. D: Appl. Phys., 2001, 34, 667–672.
60. Y. Yang, J. H. Jung, B. K. Yun, F. Zhang, K. C. Pradel, W. Guo and Z. L. Wang, Adv. Mater., 2012, 24, 5357–5362.
61. Y. Yang, W. Guo, K. C. Pradel, G. Zhu, Y. Zhou, Y. Zhang, Y. Hu, L. Lin and Z. L. Wang, Nano Lett, 2012, 12, 2833–2838.
62. Q. Leng, L. Chen, H. Guo, J. Liu, G. Liu, C. Hu and Y. Xi, J. Mater. Chem. A, 2014, 1–3.
63. D. Isakov, E. de M. Gomes, B. Almeida, A. L. Kholkin, P. Zelenovskiy, M. Neradovskiy and V. Y. Shur, Appl. Phys. Lett., 2014, 104(7), 032907.
64. T. Park, J. Na, B. Kim, Y. Kim, H. Shin and E. Kim, ACS Nano, 2015, 9, 11830–11839.
65. S. K. Ghosh, M. Xie, C. R. Bowen, P. R. Davies, D. J. Morgan and D. Mandal, Sci. Rep., 2017, 7(7), 16703.
66. P. Sadhukhan, S. Q. Wu, J. I. Long, T. Nakanishi, S. Kanegawa, K. Gao, K. Yamamoto, H. Okajima, A. Sakamoto, M. L. Baker, T. Kroll, D. Sokaras, A. Okazawa, N. Kojima, Y. Shiota, K. Yoshizawa and O. Sato, Nat Commun, 2021, 12(7), 4836.
67. Deepak, Z. Mallick, U. Sarkar, D. Mandal and R. K. Roy, Chem. Mater., 2023, 35, 3316–3328.
68. R. Haldar, A. Kumar, D. Mandal and M. Shanmugam, Mater. Horiz., 2024, 11(7), 454.

---

Footnote

† Authors contributed equally to this work.

---