Materials Science: Advances in Textile Energy Harvesting

What you’ll learn:

Why creating textiles that are suitable for energy harvesting is difficult.
Two very different approaches to addressing the challenge.
The results achieved and remaining obstacles.

Energy harvesting and scavenging are attractive power-sourcing techniques in many ways. First, it can work where conventional sources are impractical, Secondly, it has that “something for (almost) nothing” aura, which gives it a virtuous vibe. In many cases, energy harvesting resolves circuit-power problems that would otherwise be difficult to close.

While some harvesting arrangements rely on fairly complicated transducers and associated circuitry, others try to leverage items in the “everyday” around us. Those approaches often get the most attention, as they apparently can generate something useful (power) out of ordinary things you have or do anyway.

That seems to be the case with energy harvesting via thermoelectric generators (TEGs) using engineered fabrics. After all, everyone has a shirt on, right? If you could also get that shirt to provide some power, it could be a convenient, portable, local, and always-ready source.

Recognizing the attractiveness of fabric-based harvesting, many university research teams are looking at developing suitable materials and evaluating their effectiveness. Two recent examples demonstrate different approaches of fabric-driven harvesting.

KAIST’s Conductive Polymers

At the Korea Advanced Institute of Science & Technology (KAIST) in South Korea, a team of researchers developed a thermoelectric material that can be used in wearable devices, such as smart clothing, while maintaining stable thermal energy performance even in extreme environments. They claim to have resolved the longstanding dilemma associated with thermoelectric materials by establishing a balance between achieving good performance and avoiding the mechanical inflexibility and brittleness of thermoelectric materials.

The thermoelectric device generates electricity using temperature differences. If clothes are made with fiber-type thermoelectric devices, electricity can be generated from body temperature to operate other electronic devices.

^{>>Check out this TechXchange for similar articles and videos}

Power

TechXchange: Energy Harvesting

From triboelectricity to rust, these articles highlight multiple aspects of energy harvesting.

Here’s the problem, though: most heat sources are curved, such as the human body, vehicle exhaust pipes, and cooling fins. Flexible thermoelectric materials using existing polymer binders can be applied to surfaces of various shapes, but their performance is limited due to the low electrical conductivity and high thermal resistance of the polymer. At the same time, inorganic thermoelectric materials based on ceramic materials have reasonably good thermoelectric performance, but they’re fragile and difficult to produce in curved shapes.

The KAIST team noted that existing flexible thermoelectric materials contain polymer additives, but the inorganic thermoelectric material developed by the research team isn’t flexible. They overcame these limitations by twisting nanoribbons instead of additives to produce a thread-shaped thermoelectric material. Inspired by the flexibility of inorganic nanoribbons, the research team used a nanomold-based electron-beam-deposition technique to continuously deposit nanoribbons and then twisted them into that thread shape to create bismuth-telluride (Bi₂Te₃) inorganic thermoelectric fibers (Fig. 1).

1. Roadmap for imparting flexibility to brittle materials: (a) Flexibility of different materials. Ceramics and TEs are representative brittle materials. (b) Two representative methods to overcome poor mechanical properties of brittle materials. (c) Comparison between nanofilm and nanoribbon using computational methods. In the case of twisting, the nanofilm exhibits extremely high residual stress compared to the nanoribbon. (d) Twisting of nanoribbon yarn for the preparation of shape‐conformable TE yarn. (e) Digital photograph of Bi₂Te₃ yarn rolled along a metal tube. (f,g) Digital photograph (f) and SEM image (g) of as‐fabricated TE yarn.

These inorganic thermoelectric fibers have higher bending strength than existing thermoelectric materials. They showed almost no change in electrical properties even after repeated bending and tensile tests.

The resulting Bi₂Te₃ yarn, with a Seebeck coefficient of −126.6 µV/K, exhibited excellent deformability, enduring extreme bending curvatures (down to 0.5 mm), and tensile strains of ≈5% through over 1,000 cycles without significant resistance change. This allowed the yarn to be seamlessly integrated into various applications—wound around metallic pipes, embedded within life jackets, or woven into garments—demonstrating its adaptability and durability (Fig. 2).

2. Prospective applications for leveraging the intrinsic flexibility of TE yarns: (a) A schematic illustration depicting the energy-harvesting scenario exploiting the temperature differential between the hot fluid inside a pipe and the colder external air. (b) Photographic images showcasing Bi₂Te₃ yarn suspended and coiled around metallic pipes of various dimensions and shapes. (c) A schematic illustration depicting the energy-harvesting scenario leveraging the temperature contrast between a warm human body and cold seawater. (d) Photographic images demonstrating Bi₂Te₃ yarn integrated into a life jacket through knitting. (e) A schematic illustration illustrating the energy-harvesting scenario utilizing the temperature disparity between a warm human body and the surrounding cold air. (f) Photographic images exhibiting Bi2Te3 yarn incorporated into a sweater through knitting. (g) Digital photograph of the wearable TE device with the Bi₂Te₃ yarn. (h) Time‐dependent open-circuit voltage (*V_OC*) of the wearable TE generator in (g). Red and blue arrows indicate the points where the TE generator is attached to and detached from the hot surface, respectively. (i) Photographic image of a 4‐pair TE generator consisting of Bi₂Te₃ yarns and tinned copper wires. (j) Photographic images showing TE voltage generation upon heating with a heat gun.

A simple four-pair thermoelectric generator comprising Bi₂Te₃ yarns and metallic wires generated a maximum output voltage of 67.4 mV.

Full details are in their paper “Flexible All-Inorganic Thermoelectric Yarns” published in Advanced Materials, as well as its exhaustive Supporting Information file.

Chalmers University of Technology’s Coated Silk

A very different fabric-related approach was pursued by a team the Chalmers University of Technology in Sweden. They used ordinary silk thread coated with a conductive plastic material that shows promising properties for turning textiles into TEG devices.

In their previous research on this topic, the thread contained metals to maintain its stability in contact with air. However, recent advances allowed them to manufacture thread with only organic (carbon-based) polymers.

Their new type of thread features enhanced electrical conductivity and stability. It has outstanding performance stability in contact with air, while at the same time having a very good ability to conduct electricity.

The conducting polymers—formally, it’s an n-type polymer poly(benzodifurandione) known as PBFDO—have a chemical structure that allows them to conduct electricity similar to doped silicon. At the same time, they have the physical properties of plastic materials which makes them flexible. Further, these polymers don't need any of the rare earth metals commonly used in electronics.

The Chalmers team used several approaches to test their fibers as TEGs in a fabric scenario. They fabricated a textile thermocouple by hand-stitching a button using conducting yarn onto three layers of a felted wool fabric (Fig. 3).

3. Performance of the thermoelectric button: (a) Schematic representation of the thermocouple fabricated by hand-stitching a button onto a wool fabric with the n-type leg composed of PBFDO coated yarn and the p-type leg of PEDOT:PSS coated yarn. (b) Voltage V (left) and output power *P_out* (right) of the thermoelectric button as a function of current I for different temperature differences ΔT = *T_hot* − *T_cold*, where *T_hot* and *T_cold* are the temperatures of the hot plate and the cooler measured with a pair of thermocouples. (c) Open-circuit voltage *V_OC* and (d) maximum output power *P_max* as a function of ΔT recorded immediately after device fabrication or after storage at ambient conditions for 1, 11, or 41 weeks (circles represent experimental values; solid and dashed lines calculated assuming that Δ*T_tc* = ΔT and Δ*T_tc* = 0.75 · ΔT, respectively).

They decided to use a button to demonstrate the ease of handling and robustness of the n-type yarn. Furthermore, the button increases the length of the thermoelectric legs, thereby enhancing the thermal gradient experienced by the thermocouple.

The n- and p-type legs were constructed with PBFDO and PEDOT:PSS coated silk yarn, respectively. The PEDOT:PSS—Poly(3,4-ethylenedioxythiophene) polystyrene sulfonate—is a widely used composite material. PEDOT (the conductive polymer) provides electrical conductivity, and PSS (polystyrene sulfonate) acts as a counter-ion to balance the charge and improve the water solubility and processability of PEDOT. The thermoelectric properties of the n- and p-type yarns closely match, except for the opposite sign of their Seebeck coefficient.

To assess the long-term stability of the thermoelectric button, they repeatedly measured its performance. Both open-circuit voltage (V_OC) and maximum power (P_max) didn’t markedly change despite 40 weeks of storage in ambient conditions. In a further set of experiments, they fabricated a textile TEG, with PBFDO and PEDOT:PSS-coated silk yarns hand-sewn through six layers of felt wool forming eight n-legs and eight p-legs (Fig. 4).

4. Performance of the thermoelectric generator composed of eight thermocouples: (a) Schematic and photograph of the thermoelectric generator. (b) Open-circuit voltage *V_OC* as a function of ∆T (circles represent experimental values; black and grey lines calculated assuming that Δ*T_tc* = ∆T and Δ*T_tc* = 0.83 · ∆T, respectively). (c) Output power *P_out* of the thermopile as a function of current I for different temperature differences ∆T (circles represent experimental values, lines represent the results from the simulation at different ∆T considering a thermal contact resistance of 97 K cm²/W). (d) Predicted *P_max* (right) and *V_OC* (left) generated from the thermoelectric generator at ∆T = 70 K and thermal contact resistance of 97 K cm²/W as a function of thermoelectric leg length (*L_l*) (vertical grey line at *L_l* = 8 mm represents the actual *L_l* of the device). The electrical contact resistance *L_l* per thermocouple was 1 Ω (triangles), 3 Ω (circles), 9.1 Ω (squares) (which corresponds to the actual *L_l* of the device at ∆T = 70 K), and the thread count *N_t* was set to be 90. (e) Predicted *P_max* (right) and *V_OC* (left) generated from the thermoelectric generator at ∆T = 70 K and thermal contact resistance of 97 K cm²/W as a function of *L_l* (vertical grey line at *L_l* = 8 mm represents the actual *L_l* of the device). The electrical contact resistance *R_C* per thermocouple was 9.1 Ω (which corresponds to the actual *R_C* of the device at ∆T = 70 K), and the thread count *N_t* was 90 (squares), 180 (circles), and 270 (triangles).

In a final set of experiments, they monitored the thermoelectric performance of the textile device with repeated bending around a coffee mug with a diameter of 10 cm. The internal resistance remained largely unaffected by bending and unbending cycles, with variations of less than 2.5% compared to the initial value after 12 cycles.

They also measured the thermoelectric performance of the bent device and observed that repeated bending didn’t affect the thermoelectric performance, which continued to function as before the mechanical deformation was applied.

For any fabric, questions regarding the impact of wash cycles are always relevant. Here the news is less favorable, as seven machine washing cycles at 20°C resulted in an increase in electrical resistance by a factor of 3, suggesting that mechanical wear limits the stability.

The work is published in a detailed paper “Poly(benzodifurandione) Coated Silk Yarn for Thermoelectric Textiles” in Advanced Materials; there’s also a Supporting Information posting.

What’s Next for Textile Harvesting?

While these fabric-centric TEGs are interesting, the reality is that their output power is quite low even with moderate-to-large temperature differences, as is their power density per square unit area. There are also practical issues of making a reliable, rugged connection to the fabric to draw on any harvested power—one that must also withstand laundering and other forms of “abuse.”

Furthermore, to what extent will an industry standard (IPC-8921, Requirements for Woven and Knitted Electronics Textiles (E-Textiles) Integrated with Conductive Fibers, Conductive Yarns and/or Wires) help or perhaps even delay progress?

These sorts of research-driven breakthroughs in the lab environment often aren’t a success in the broader market due to issues related to scaled-up performance, manufacturing, reliability, or cost. This is analogous to the many real-world problems encountered by nearly all heralded battery-related advances we see touted as “revolutionary” and “breakthrough.”

Nonetheless, it’s compelling to watch how materials science is advancing the TEG harvesting prospects and creating potential harvesting options. It will be interesting to see if fabric-based energy-harvesting TEGs have a viable commercial future, or if they’re destined to be restricted to lab curiosities.