Flexible triboelectric nanogenerators using transparent copper nanowire electrodes: energy harvesting, sensing human activities and material recognition

Triboelectric nanogenerators (TENGs) have emerged as a promising green technology to efficiently harvest otherwise wasted mechanical energy from the environment and human activities. However, cost-effective and reliably performing TENGs require rational integration of triboelectric materials, spacers, and electrodes. The present work reports for the first time the use of oxydation-resistant pure copper nanowires (CuNWs) as an electrode to develop a flexible, and inexpensive TENG through a potentially scalable approach involving vacuum filtration and lactic acid treatment. A ∼6 cm2 device yields a remarkable open circuit voltage (Voc) of 200 V and power density of 10.67 W m−2 under human finger tapping. The device is robust, flexible and noncytotoxic as assessed by stretching/bending maneuvers, corrosion tests, continuous operation for 8000 cycles, and biocompatibility tests using human fibroblast cells. The device can power 115 light emitting diodes (LEDs) and a digital calculator; sense bending and motion from the human hand; and transmit Morse code signals. The robustness, flexibility, transparency, and non-cytotoxicity of the device render it particularly promising for a wide range of energy harvesting and advanced healthcare applications, such as sensorised smart gloves for tactile sensing, material identification and safer surgical intervention.

Copper and silver based nanostructured electrodes have been implemented in previous TENGs by using various combinations of PDMS and additives such as graphene under both single and double electrode mode. However., a comparative analysis of parameters such as output voltage, current, power density and electrode surface area shows that our copper nanowire based TENG exhibits highest power density obtained from human finger tapping in single electrode mode with an active area of 6 cm 2 (see table S1). In contrast to previously reported methods, our approach utilises oxygen stable copper nanowires (obtained after treatment with lactic acid) without any additive or additional processing steps which benefits from the high charge collecting capacity of the pristine metallic copper. This green approach has inherent advantages of facile synthesis, high conductivity and stability which is reflected in the TENG performance. Finally, although it was not investigated, the rough microstructure of CuNW may generate additional charge at the PDMS-Cu interface during mechanical impact augmenting TENG output.

Performance of transparent TENG
The transparent TENG was fabricated simply by using a diluted concentration of CuNW (10 times less than the original TENG). Understandably, the electrical output and performance was low as represented in Figure S1 due the less amount of copper present in the electrode. While the power density may not be high enough for energy harvesting applications, these can be potential candidate for implantable TENG-based sensors as less concentration of copper translates to lower cytotoxicity in biological tissues as observed in our previous study 9 .

Substrate versatility of TENG
The TENG can be fabricated using any adhesive flexible substrate which can be tailored for specific applications. For example, different commercial adhesive tapes were used as a substrate, on which CuNW was deposited to make TENGs of variable robustness and transparency. This allows for a wide range of flexible or even stretchable substrates that can used to make conformable TENGs for energy harvesting and sensing from human body movements. Figure S2: Flexible TENGs fabricated based on different adhesive substrate-a) insulating tape, b) masking tape, c) Gorilla tape, d) double sided tape and e) and f) transparent Kapton tape.

Output from different triboelectric materials
To further highlight the versatility of the TENG, voltage output was measured against some polymeric materials which are conventionally used in packaging, textile and healthcare applications. The TENG was tapped with nylon based textiles, rubber surgical gloves, and PES filter membrane to show the potential of energy harvesting from commonly used materials.

Code for Morse code generation using TENG
The following algorithm was used to convert the voltage output from the TENG during tapping with finger to generate and transmit Morse codes on a laptop. Each alphabet was assigned as a 'dot' or 'dash' (following Morse code) value based on the intensity of the output voltage. The TENG successfully generated voltage outputs consistent with the assigned alphabets to form common words on the laptop screen. #%% # Find number of peaks in signal from pathlib import Path import numpy as np import pandas as pd import matplotlib.pyplot as plt from scipy.signal import find_peaks

Voltage pattern from holding different objects
The sensor mounted glove was used to hold and interact with different objects which produced voltage patterns and peak intensities based on respective triboelectric pairs (metal-Ecoflex, plastic-Ecoflex, glass-Ecoflex). Detailed analysis of these voltage outputs (frequency, intensity and overall pattern) may be beneficial to extract information about different human-device interactions and potentially aid in object recognition and robotic surgery in future. Figure S5: Open circuit voltage (V oc ) obtained from TENG sensorized glove upon grasping different objects in single electrode mode.

Force calibration of TENG
Force was calibrated by applying a known impact force on the TENG surface and then simultaneously recording the voltage output (by digital oscilloscope and force (by using a force plate-see section 2.6). The impact force was applied 20 times and varied from 3-30 N. Figure  S6 shows the plot of Voltage output vs applied force. The output voltage acquired from the TENG increased linearly with force showing a working range of 5-30 N. Figure S6: Voltage (V oc ) vs Force calibration curve using TENG. The TENG was tapped with different controlled forces (force was measured directly by placing the device on a force plate) and the corresponding output voltage was recorded by an oscilloscope.

Material identification with TENG mounted glove
An attempt has been made to use TENG sensor mounted glove as a simple analytical tool to identify material characteristics by exploiting differential triboelectric charge generation between Ecoflex and some commonly used known materials. The voltage output obtained after tapping on each material is then analysed based on two identifying parameters like 'polarity' and 'intensity' (see section 3) and finally compared to an unknown material. Material is then identified based on minimal difference between the above mentioned parameters. The material polyimide was successfully (98% success rate) identified by comparing with 7 different materials based on output voltage and polarity. Although this is a crude and qualitative method, testing with wider range of materials followed by careful analysis of the voltage/current outputs using stringent conditions and comparing them in the triboelectric series may open up possibilities to develop touch based material identification strategies in future 10 . Video S10: TENG detecting finger joint movement.
Video S11: TENG based sensorised surgical glove showing variation of voltage output from tapping on different surfaces.