High performance temperature difference triboelectric nanogenerator

Triboelectric nanogenerators (TENGs) convert mechanical energy into electricity via contact electrification and electrostatic induction. Enhancing TENG output has been pursued through material selection, structural optimization, prior-charge injection, operating in vacuum, and ferroelectric/dielectric engineering. However, many applications require operation at elevated temperatures (e.g., engines), where performance degrades above ~260 K due to increased thermionic emission and reduced charge storage on friction layers. Temperature between friction layers influences electron transfer: higher temperature promotes electron transfer from hotter to cooler materials but also increases thermionic emission from the cooler layer as it warms through heat exchange. Prior nanoscale AFM studies indicated temperature difference can boost charge transfer within a certain ΔT window. This work investigates whether and how a controlled temperature difference (ΔT) between friction layers can enhance macroscopic TENG outputs, identifies an optimal ΔT, and demonstrates a temperature-difference TENG (TDNG) with controllable layer temperatures for high-temperature environments.

Previous TENG improvements include: pumping/self-charge systems achieving 25.2–28.8 µA cm−2 in air; vacuum operation reaching 57 µA cm−2; prior charge injection to 90 µA cm−2; electric double-layer designs up to 305 µA cm−2; ferroelectric/dielectric permittivity modulation pushing to 350 µA cm−2. Temperature effects: output degrades at higher working temperatures due to thermionic emission and reduced charge retention. Electron thermionic emission models have explained temperature-dependent output. Nanoscale AFM studies (metal–dielectric) showed ΔT facilitates electron transfer from hotter to cooler surfaces and can raise charge density within a certain ΔT range. Practical constraints include heat exchange during contact–separation causing the cooler layer to warm, diminishing gains. This background motivates exploiting ΔT in a macroscopic TENG via thermal management.

Theory and simulation: An electron-cloud–potential-well model describes triboelectrification between hotter and cooler friction layers under a temperature difference. Two competing effects are considered: (i) increased electron transfer from the hotter to the cooler layer as the hotter layer’s electron energy levels rise with temperature; (ii) enhanced thermionic emission leading to charge dissipation from the cooler layer as its temperature rises due to heat exchange. COMSOL simulations incorporate (a) a linear increase of the cooler layer temperature with that of the hotter layer and (b) empirical/modified models relating surface charge density to cooler temperature (σ = −C1 T1 + C2) and thermionic emission C = e^(−S/T0), combining to predict σ(ΔT), surface potential, and transferred charge trends. The model also analyzes the role of the cooler layer’s thermal conductivity on optimal ΔT and charge density. A classical electrodynamics expression relates output voltage V(t) to σ(ΔT), dielectric thickness, permittivity, separation distance x(t), and load resistance. Device design and fabrication: A temperature-difference TENG (TDNG) comprises a hotter part and a cooler part separated by an air gap. Hotter part friction layer: 20-µm-thick Al foil etched chemically to form ~100 nm-scale nanostructures. Cooler part friction layer: 100-µm-thick Kapton film etched by RIE to create ~150 nm wide, ~100 nm high nanopillars. Electrodes: Cr/Ag sputtered (4.8 cm × 4.8 cm) on Kapton. Heating/cooling: thermostat heater for the hotter part and a water-cooling system for the cooler part to independently control layer temperatures. Air gap: ~5 cm. Nanostructuring increases effective surface areas (~1.87× for Al and ~1.96× for Kapton). Additional TDNG variants employed different friction pairs: Al–Kapton, PA-6–Kapton, Cu–Kapton, Fe–Kapton, and Al–PTFE. Experimental setup and measurements: A linear motor (0.7 Hz baseline; other frequencies/velocities also tested) drives contact–separation. Outputs measured: open-circuit voltage (with SR560 and voltage divider), short-circuit current (SR570), transferred charge per cycle (CQC), surface potential (Kelvin probe), and thermally stimulated discharge (TSD) current of Kapton to evaluate stored charges. Load-dependence characterized across resistances to determine maximum power. Parameter sweeps: ΔT from 0 to 219 K at fixed cooler reference (~299 K), frequency (0.18–0.53 Hz), contact–separation velocity (30–1286 mm s−1), and external load. For high-performance demonstrations, an Al–PTFE TDNG operated at ΔT ≈ 90 K with average applied pressure ~94.5 kPa and high contact–separation velocity. Wind-driven TDNG: substituted 15 cm × 7 cm Al/Kapton layers; hotter Al attached to heater, Kapton flag free in airflow (air gap ~1 cm) to harvest wind-induced motion over hot surfaces; outputs rectified for capacitor charging and LED/sensor demonstrations.

• Simulations and theory predict a non-monotonic dependence of transferred charge density and surface potential on ΔT due to competition between enhanced electron transfer (hot→cool) and thermionic discharge from the cooler layer; an optimal ΔT exists. • Experimental Al–Kapton TDNG at ~0.7 Hz: as ΔT increases from 0 to 219 K, output voltage and current first increase then decrease. At optimal ΔT ≈ 145 K: Voc ≈ 858 V, Isc ≈ 20 µA, surface charge density ≈ 58.8 µC m−2, and output power ≈ 206.7 µW. • Relative improvements (ΔT = 145 K vs 0 K): Voc ×2.7 (≈315→858 V), Isc ×2.2 (≈9.0→20 µA), σ ×3.0 (≈19.6→58.8 µC m−2). Power improvement depends on baseline: under load-sweep, maximum power increases from ≈42.2 µW to ≈206.7 µW (×4.9 at 3 MΩ); in initial conditions referenced in the abstract (~69 µW→206.7 µW), ≈3×. • Transferred charge per cycle (CQC) peaks at ΔT ≈ 145 K with 147 nC (≈58.8 µC m−2), about 3× that at ΔT = 0 K. TSD measurements show maximum integrated discharge near ΔT ≈ 155 K. Surface potential of Kapton reaches −112 V at ΔT ≈ 145 K, confirming optimal ΔT for maximizing stored charges. • Material dependence: All tested friction pairs benefit from ΔT. At optimal ΔT, Voc and Isc enhancement factors respectively: Al–Kapton (2.7×, 2.2×; optimal ΔT ~145 K), PA‑6–Kapton (3.2×, 2.0×; ~144 K), Cu–Kapton (3.0×, 2.6×; ~145 K), Fe–Kapton (3.9×, 10×; ~143 K), Al–PTFE (2.7×, 1.7×; ~90 K). Optimal ΔT depends on cooler-layer material and its thermal conductivity. • High-performance Al–PTFE TDNG (ΔT ≈ 90 K, f ≈ 0.8 Hz, average pressure ~94.5 kPa, velocity ~1286 mm s−1): peak Voc up to 2.11 kV (average ≈1.80 kV); peak short-circuit current density up to 670 µA cm−2; average peak current density ≈480.3 µA cm−2; across 110 cycles (10 tests), average current density ≈443 ± 46.6 µA cm−2, exceeding the prior record (350 µA cm−2) by ≈26.6%. • Load characteristics: Maximum output power at ~3 MΩ; with ΔT = 145 K, power increases to ≈206.7 µW vs ≈42.2 µW at ΔT = 0 K (×4.9). • Wind-driven TDNG on hot surface: at wind speed ~5 m s−1, maximum current ≈123.3 µA at ΔT ≈ 38.7 K; transferred charge ≈185.7 µC. A 22 µF capacitor charges to ~7.5 V within 60 s at ΔT ≈ 41 K (faster and higher than at ΔT = 0 or 21 K). Demonstrated lighting 955 LEDs and powering a temperature–humidity sensor after ~73 s charging at ΔT ≈ 102 K and wind ~8.3 m s−1.

The results validate that imposing a controlled temperature difference between friction layers can counteract the typical high-temperature degradation of TENGs by enhancing electron transfer from hotter to cooler surfaces, while acknowledging thermionic discharge as ΔT increases. The non-monotonic behavior with a clear optimal ΔT arises from this tradeoff. The optimal ΔT depends on the cooler material’s thermal properties; higher thermal conductivity can shift and raise the optimal ΔT and achievable surface charge density. The approach generalizes across multiple material pairs, enabling substantial gains in voltage, current, charge density, and output power without complex structural modifications or prior-charge treatments. High current densities exceeding previous records are achieved by selecting favorable material combinations (e.g., Al–PTFE) and operating conditions, demonstrating practical relevance for harvesting mechanical energy in high-temperature environments. Wind-driven demonstrations further show that ΔT-assisted TENGs can utilize both thermal gradients and ambient motion to power devices and indicators.

This work establishes temperature difference as a powerful lever to boost TENG performance at elevated temperatures. A TDNG with independent thermal control of friction layers demonstrates that an optimal ΔT (~145 K for Al–Kapton; ~90 K for Al–PTFE) maximizes charge transfer and output by balancing enhanced electron transfer against thermionic discharge. Measured outputs include Voc up to ~2.1 kV and current densities averaging ~443 µA cm−2 (peaks to 670 µA cm−2), surpassing prior records. The strategy is broadly applicable across material pairs and compatible with practical scenarios, as shown by a wind-driven TDNG powering hundreds of LEDs and a sensor on hot surfaces. Future work can explore material and thermal engineering to maintain optimal ΔT under dynamic conditions, integrate passive/active thermal management to minimize cooler-layer heating during operation, and scale device architectures for robust deployment in harsh, high-temperature environments.

Maintaining a stable temperature difference in practical operation is challenging due to heat exchange through air and contact, which warms the cooler friction layer and reduces performance. Optimal ΔT and gains are material dependent, influenced by the cooler layer’s thermal conductivity. Laboratory demonstrations rely on external heating/cooling systems and controlled mechanical actuation, which may not directly translate to all real-world conditions. Performance metrics also depend on driving frequency, contact pressure, and separation velocity.