Wearable thermoelectric generator for harvesting human body heat energy

When the number of the thermocouples is n, the Seebeck coefficient of the used thermoelectric materials is α, and the temperature difference between hot and cold junctions is ΔT_TEG ; an output voltage V is generated by the Seebeck effect:

$$\begin{eqnarray}&&V=n\alpha \Delta {{T}_{TEG}}\end{eqnarray} \tag{ 1 }$$

The output power P is dependent on the internal electrical resistance R_G and the external load R_L :

$$\begin{eqnarray}&&P=\;\frac{{{V}^{2}}}{{{({{R}_{G}}+{{R}_{L}})}^{2}}}\cdot {{R}_{L}}\end{eqnarray} \tag{ 2 }$$

When R_L is equal to the internal electrical resistance R_G , which provides optimal impedance matching, the maximum output power P₀ is achieved:

$$\begin{eqnarray}&&{{P}_{0}}=\;\frac{{{V}^{2}}}{4{{R}_{G}}}\end{eqnarray} \tag{ 3 }$$

The equivalent thermal circuit of the TEG on the skin is depicted in figure 1. The core temperature of the human body is about 36 °C, and the ambient temperature is from −10 to 40 °C, depending on the environment. A TEG is placed in contact with the skin, and another of its surfaces is open to the ambient air. The temperature difference of the TEG (ΔT_TEG ), which is defined as the difference between the temperature of the outer surface T_surface and the temperature of the skin T_skin, can be expressed as follows:

$$\begin{eqnarray}&&{{R}_{TEG}}=\;\frac{{{R}_{TE\,}}\cdot \;{{R}_{{\rm fabric}}}}{{{R}_{TE}}\;+\;{{R}_{{\rm fabric}}}}\end{eqnarray} \tag{ 4 }$$

$$\begin{eqnarray}&&\Delta {{T}_{TEG}}=\;\frac{{{R}_{TEG}}}{{{R}_{{\rm air}}}+{{R}_{TEG}}+{{R}_{{\rm body}}}}\;\cdot \Delta T\end{eqnarray} \tag{ 5 }$$

Zoom In Zoom Out Reset image size

The output voltage is proportional to ΔT_TEG , as shown in equation (1). The ambient temperature has the greatest effect on the ΔT_TEG . The high thermal resistivities of the human skin and of the ambient air affect the total thermal resistance in the thermal circuit. Therefore, the temperature difference across the TEG is significantly lower than the measured temperature difference between the air and the body. Each individual thermocouple has a low thermal resistance; the thermal resistance of a single thermocouple is not large enough to cause a large temperature difference between the surfaces of the TEG. Furthermore, the parallel arrangement of the thermocouples makes the total thermal resistance even smaller [9]. Consequently, even a large temperature difference ΔT will result in a relatively small temperature difference between the hot and cold junctions because the thermal resistance of the TEG R_TEG is so small, as can be seen from equation (5). However, environmental changes can cause a more active heat exchange on the surface. The thermal resistance can be reduced by air convection caused by wind or other naturally occurring air currents or by human movement. Thus, the performance of the thermoelectric device can vary depending on the situation.

Figure 2 shows the concept views of the proposed wearable TEG. As shown in the figure, the form of the device allows it to be easily embedded in clothing. Thermoelectric columns of p- and n-type material are located in the polymer-based fabric. Each thermocouple is electrically connected by a conductive fabric fiber. A polyimide film is attached to the bottom of the TEG and prevents physical and electrical contact of the TEG with the skin. The polyimide film was chosen because of its superior thermal and electrical resistance [10].

Zoom In Zoom Out Reset image size

Geometrical details of the prototype are shown in figure 3. The prototype with 12 integrated thermocouples has an area of 6 × 25 mm². The element radius of the thermocouple has a diameter of 4 mm with a thickness of 0.6 mm, and the spacing between the elements is 4 mm. The attached polyimide film thickness is 100 μm. The geometries of each thermocouple were determined with reference to the optimization design research [11]. However, the arbitrary selected design has not been optimized for area and density. There are restrictions on the design. Increasing the number of thermocouple elements per unit area decreases the flexibility of the device. Thus, sufficient spacing between the elements is required.

Zoom In Zoom Out Reset image size

We have previously proposed a flexible TEG using PDMS (polydimethylsiloxane) for human body heat. The device had a good flexibility, but it was not good enough for wearing because of a thick PDMS substrate [12]. In this work, we demonstrate wearable TEG and better performance for low temperature energy harvesting. The thermocouple of the prototype consisted of Bi_0.5Sb_1.5Te₃ (p-type) and Bi₂Se_0.3Te_2.7 (n-type). These powdered materials (325 mesh), purchased from RND Korea Comp., show better thermoelectric properties than the Bi₂Te₃ used in our previous work [12]. Liquid binder is used to solidify the powders; it is a refractory ceramic colloid in water (Durabond 950, IS-Rayfast Ltd). The binder has a low curing temperature and can be used with a variety of metallic powders. After curing, it has excellent electrical properties and durability.

To make printable ink, the Bi_0.5Sb_1.5Te₃ and Bi₂Se_0.3Te_2.7 powders were suspended in the adhesive binder solution. 78–82 wt% of powder and 18–22 wt% of binder were mixed with a vortex mixer. A dispenser was used to fill the holes in the fabric with the thermoelectric ink. Dispenser printing is very simple. The viscous printable ink is injected using a syringe that is controlled by the program. The dispenser printing has a low temperature fabrication step. Feature sizes can be adjustable depending on process parameters such as shot pressure, tip size, ink rheology and shot spacing [13]. The dispensed ink was dried at room temperature for 24 h, followed by curing at 100 °C in a vacuum oven for 2 h. Final curing at a higher temperature (200–300 °C) is required. However, we did not cure at such a higher temperature in order to prevent damage on the fabric. The synthesized materials are in the form of a uniform and dense matrix. Table 1 shows the Seebeck coefficient of the fabricated thermoelectric materials from measured data. The results were not excellent, but the dispensing process provides a fast and inexpensive way to make a solid material.

A polymer-based fabric was chosen as a substrate because it provides a flexible and lightweight material. A conductive thread was used for making electrical connections. Thread that contains silver has good electrical properties and flexibility (Amosilver, Amo-wire Co.). Using these fabric materials, which are soft and lightweight, ensures that the human wearer can move comfortably in them [10].

A schematic of the sewing process is shown in figure 4. This process is needed for fixing conductive fiber to the prepared substrate fabric. First, the conductive thread was sewn on the fabric. The thread was cut in accordance with the p- and n-sections. Then, the process was repeated on the other side in the same way. The polyimide film was attached to the bottom of the fabric layer. Next, the prepared printable thermoelectric ink was dispensed into the window of the fabric. Finally, the device was cured at 100 °C for 2 h.

Zoom In Zoom Out Reset image size

A top view (figure 5(a)) and a close-up view (figure 5(b)) of the fabricated device are shown. As shown in figure 5(c), the TEG is highly flexible and durable. After several bending tests, there was no further change to the electrical properties of the TEG. We confirmed that the TEG is strong enough for flexible applications. The total internal resistance of the 12 thermocouples is about 902 Ω.

Zoom In Zoom Out Reset image size

The wearable TEG was tested and characterized with the measurement setup. For this experiment, the setup consisted of a hot plate and an Al heat spreader (see the figure 7 inset). To reduce the thermal contact resistance, thermal paste was applied to the contact surface between the TEG and the copper plate.

Figure 6 shows the 12-couple TEG characteristic curves for the prototype at ΔT = 20 K. The power output was measured at various load resistances. The optimal output of the TEG occurred when the load resistance matched the internal resistance of about 900 Ω. At matched load resistance, the optimal power output is approximately 224 nW at 15.8 μA and 14.2 mV. Figure 7 shows the maximum voltage and power at matched load resistance that was measured with the variation of the temperature difference (ΔT_TEG ). The temperature difference was considered as the measured data at the TEG surface of the top and bottom. The graph clearly shows the linearly increasing voltage with the increasing temperature difference. At the temperature difference of 19.2 K, the TEG generated a voltage of 14.1 mV and a power of 221 nW. A thermopower of 0.72 mV K⁻¹ could be attained with a single device with only 12 thermocouples.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

It was possible to integrate the TEG in the fabric of a shirt, as shown in figure 8. The fabricated device was worn on the wearer's chest; the chest is a large area and is the warmest part of the body. The measured skin temperature was about 32 °C, and the ambient temperatures for the cases of cold and hot environments were 5 °C and 25 °C, respectively. Table 2 shows the results of harvesting energy from the human body. All of the measurements have been carried out under outdoor conditions in a natural convection situation in which the wearer did not move. In the warm ambient air of 25 °C, the device generated less power (8.1 nW), but the output power of the TEG increased considerably, to as high as 146.8 nW, in the cold ambient air (5 °C).

Zoom In Zoom Out Reset image size

When the wearer is walking, a significant amount of convection occurs, and the skin surface temperature is raised by the activity of movement. Thereby the amount of harvested power increased. We measured the output voltage as a function of time for two different states of motion (figure 9). The voltage change was observed for 1 min immediately after the subject began wearing the TEG. This figure shows that the voltage decreases as a function of time. The voltage reduction occurred due to the temperature equilibrium between the hot and cold junction. The first time the TEG came in contact with the skin, the temperature difference was the highest. But after the heat was transferred, the temperature difference decreased. After that, the voltage was maintained at a constant level by thermal equilibrium. A higher voltage was maintained while walking than in the stationary state. These results were maintained when running the experiment 10 times; averaging these experimental results yielded voltages of about 2.7 mV (stopped state) and 4.8 mV (walking state). The walking speed was approximately 1 m s⁻¹, and the ambient temperature was measured to be 25 °C.

Zoom In Zoom Out Reset image size