Self-rechargeable cardiac pacemaker system with triboelectric nanogenerators

The study addresses the challenge of providing sustainable power to implantable medical devices, which currently rely on finite-life batteries that may necessitate risky replacement surgeries. Prior biomechanical energy harvesters often require direct mechanical deformation and are difficult to integrate within rigid, hermetically sealed, titanium-packaged implants. The authors propose an inertia-driven triboelectric nanogenerator (I-TENG) that converts body motion and gravity-induced inertial forces into electrical energy while being fully enclosed in biocompatible materials. The research aims to demonstrate a compact, stacked I-TENG capable of harvesting sufficient in vivo energy to recharge a battery and to power a functional cardiac pacemaker system, thereby advancing the feasibility of self-rechargeable implantable devices.

The paper situates its contribution within multiple prior approaches to powering implants: wireless power transfer to deep tissue microimplants, and various biomechanical energy harvesters including piezoelectric and triboelectric nanogenerators for in vivo sensing, monitoring, and even biodegradable power sources. Previous TENG-based implantables typically relied on direct mechanical deformation and faced packaging/biocompatibility constraints. A prior symbiotic pacemaker concept harvested cardiac motion. The authors highlight limitations of synchronizing multiple generators and the inefficiencies introduced by rectification and energy management circuits. They emphasize the need for enclosed, biocompatible designs that can leverage inertial motion, as well as advances in surface treatments to boost triboelectric charge density and output. The references include works on wireless power, implantable/wearable TENGs, biodegradable TENGs, Li-ion charging with pulsed TENG outputs, and strategies to enhance TENG performance.

Device design and fabrication: An inertia-driven TENG (I-TENG) was designed with a freestanding unit oscillating between triboelectric layers under inertial forces and gravity. A 50 µm PFA film (10 cm × 10 cm) underwent surface modification via reactive ion etching (RIE) with 50 sccm O2 and 50 sccm Ar at 100 W, 0.2 Torr for 5 min. The PFA was then electrically polarized by corona poling: needles 1 cm above the film connected to cathode, Cu plate as ground; 15 kV applied for 15 min; then film inverted and re-poled at 15 kV for 15 min. The PFA film was attached to a Cu mass (thickness 0.8 mm; radius 1.25 cm) using carbon double-sided tape. A PVA-NH2 solution was spin-coated on an Au-deposited substrate (radius 1.5 cm). A 2 mm thick acrylic layer served as a gap between top and bottom substrates. The assembled I-TENG was encapsulated with a medical-grade biocompatible silicone (C6-540 Liquid Silicone Rubber, ~5 mm thick), chemically sterilized, and rinsed for animal experiments. Stacked architecture and synchronization: Five I-TENG units were stacked in a single package. SPICE simulations compared asynchronous versus synchronous operation; the encapsulated and commonly actuated design enabled synchronized current phases across units, allowing current superposition without added rectification per unit. Electrical characterization: A shaker applied regular vertical displacement (e.g., 3 Hz). Output voltage/current were measured across load resistances (notably ~10 MΩ) using a Tektronix DPO 3052 oscilloscope and SR570 low-noise current preamplifier. RMS power and current density versus load were obtained; stability over >30,000 cycles was evaluated. Internal energy conversion efficiency was calculated as the ratio of output electric energy (WRMS over one period) to input kinetic energy (m v^2/2), yielding ~0.235% for the five-stack. Power management and energy storage: The I-TENG output was routed through a low-pass filter to a first energy storage (parallel low-ESL capacitors, 60 µF). A PMIC transferred energy from the first storage to a larger second storage (parallel low-ESR capacitors, 150 µF), reducing voltage while conserving energy. When the second storage exceeded threshold (e.g., ~3.4 V in lab; >1.8 V in vivo demonstration), energy was transferred to a Li-ion microbattery (ML414H, 1 mAh) for charging. Kinematics modeling: The relationship among acceleration, working angle, and output was modeled. Net force on the freestanding unit comprises inertial and gravitational components. Vertical displacement dver and open-circuit voltage VOC were related to the vertical acceleration component (a(t) sinθ − g), showing that z-axis displacement drives performance while x-axis lateral motion minimally contributes to induction; displacements >4 mm yielded >80% relative power. In vivo implantation and monitoring: Packaged devices were implanted subcutaneously on the backs of adult mongrels under general anesthesia. Two devices were oriented with different normals (horizontal vs vertical relative to body axis) to assess directional sensitivity. A BLE-based wireless system monitored real-time output voltage and capacitor voltages; due to system limits, transmitted signals saturated at 3.75 V and had a 0.5 V baseline in standby. Device activity was recorded over 24 hours, including daytime activity and nighttime rest. First and second energy storage charging profiles were tracked in vivo, and battery charging events were demonstrated when thresholds were reached. Biocompatibility assessment: After implantation in contact with muscle, Masson’s trichrome staining evaluated inflammatory responses and fibrosis around the encapsulated I-TENG compared with normal muscle and a commercial medical device. Fibrosis area percentage was quantified (n=5 per group) using ImageJ; one-way ANOVA (p=0.0002) assessed differences. Pacemaker integration: A titanium-packaged self-rechargeable cardiac pacemaker integrated the stacked I-TENG, power management and battery charging, and pacing/EGM sensing electronics. The Ti shell (0.5–1.5 mm thick) ensured biocompatibility; the lead connector was encapsulated in medical-grade silicone. The system sensed atrial and ventricular EGM, supported asynchronous VOO pacing at 150 bpm, and VVI mode. In vivo, ventricular pacing parameters included 5 V, 1 ms pulse width; ventricular sensing 1.4 mV; lower rate interval 90 bpm. Bradycardia was induced via adenosine and via vagus nerve stimulation (12 mA, 1 ms, 13 Hz) to test automatic pacing response. EKG/EGM recordings validated sensing and pacing behavior.

• Five-stacked I-TENG synchronized output without per-unit rectification, amplifying current by superposition.
• Output scaling with stacking (3 Hz, ~10 MΩ load): peak voltage increased from 36 V (1 stack) to 136 V (5 stacks); peak current volume density increased from 0.4 to 2 µA/cm³.
• Maximum RMS power density: 4.9 µW/cm³ at ~10 MΩ matched load; stable operation over >30,000 cycles.
• Internal energy conversion efficiency of the five-stack: ~0.235% (output electric energy 11.43 µJ vs input kinetic energy estimate as specified).
• Kinematics dependence: Output voltage scales nearly linearly with z-axis displacement; x-axis displacement minimally affects performance; z-displacement >4 mm yields >80% relative power.
• Power management: Efficient multi-stage storage allowed stepping down high voltages to charge a 1 mAh Li-ion battery; in lab, second storage charged to ~3.4 V before battery transfer.
• In vivo performance (adult mongrels): Real-time BLE-monitored outputs up to the system cap (3.75 V), with standby 0.5 V baseline. Orientation affected activity-induced harvesting; device aligned with more frequent motion showed higher output. Over 24 hours, the system harvested substantial energy; the study reports around 144 mW of power during active periods. Both first (60 µF) and second (150 µF) storages charged in vivo; transfer to Li-ion battery occurred after second storage exceeded ~1.8 V.
• Biocompatibility: Encapsulated I-TENG elicited mild inflammatory response and prominent fibrotic capsule comparable to a commercial device; no infection observed. Quantified fibrosis area differed from normal muscle but was similar between devices (n=5, one-way ANOVA p=0.0002 for group differences).
• Pacemaker integration: Demonstrated functional self-rechargeable pacemaker in vivo. VOO mode at 150 bpm and VVI mode operation confirmed. With LRI 90 bpm, device sensed ventricular signals (~1.4 mV) and delivered 5 V, 1 ms pulses when intrinsic rate fell below threshold (e.g., after adenosine or vagus nerve stimulation reducing rate to ~80 bpm), restoring rate to 90 bpm; pacing ceased upon recovery.

The results show that an inertia-driven, fully enclosed TENG can convert body motion and gravitational inertial forces into usable electrical power for implants, overcoming limitations of prior TENGs that required direct deformation and were difficult to hermetically package. Stacking synchronized units inside a single package increased current without complex rectification, enabling efficient power extraction and storage. Laboratory tests verified power density, efficiency, and durability, while kinematic analysis clarified that ensuring sufficient z-axis displacement is key to reliable output despite multidirectional motion. Preclinical in vivo tests validated real-time energy harvesting during natural activity, successful staged energy storage, and battery charging within the body. Integrating the I-TENG with a titanium-packaged pacemaker demonstrated clinically relevant sensing and pacing modes (VOO, VVI), with autonomous intervention during induced bradycardia. Together, these findings address the research goal of achieving a self-rechargeable implantable system, suggesting viability for reducing battery replacement surgeries and enabling long-term operation of medical implants.

This work introduces a compact, stacked inertia-driven TENG (I-TENG) that is fully encapsulated in biocompatible materials and capable of harvesting biomechanical energy in vivo to recharge a battery and power a functional cardiac pacemaker. The device achieves synchronized multi-unit output, a maximum RMS power density of 4.9 µW/cm³, and robust operation, and it demonstrates effective power management and battery charging both in laboratory and preclinical animal models. The integrated pacemaker system successfully performed sensing and pacing (VOO, VVI) in vivo, responding appropriately to bradycardia. Future research should focus on improving energy conversion efficiency and output under realistic activity profiles, optimizing packaging and mechanical coupling to maximize z-axis displacement, refining low-power electronics and PMICs for higher end-to-end efficiency, long-term chronic implantation studies to assess stability and tissue response, and extending the approach to other implantable devices and multimodal health-monitoring systems.

• Performance discrepancy between laboratory and in vivo conditions due to variability and lower magnitude of available mechanical energy from animal motion.
• The BLE wireless measurement system imposed voltage range limits (saturation at ~3.75 V, 0.5 V baseline), constraining real-time signal fidelity.
• Internal energy conversion efficiency remains modest (~0.235%), indicating room for materials and structural optimization.
• Effective harvesting depends on sufficient z-axis displacement; suboptimal orientation or motion profiles may reduce output.
• Preclinical testing was limited to short-term studies in a small number of animals; long-term chronic biocompatibility, mechanical durability, and functional stability were not fully assessed.
• Mild fibrotic encapsulation was observed; while comparable to commercial devices, its long-term impact on mechanical coupling and energy transfer warrants further study.