Introduction

Fiber drawing from a preform has traditionally been exploited to produce the extended lengths of optical fibers as telecommunication waveguides1,2 to deliver enormous amounts of light data over the kilometer length scale3,4,5. Though initially developed for optical fibers, preform-to-fiber thermal drawing technique is an inherently scalable manufacturing method that produces functional fibers with the integration of a broad range of materials6,7,8, including glasses, semiconductors, metals, and polymers. It has enabled a wealth of functionalities, including material engineering9,10,11, optical sensing12,13, neural sensing14,15, thermal detection16,17, chemical sensing18,19, acoustic emission20,21, and many more22. Prospectively, such functional fibers can be a promising candidate for soft electronics23,24,25,26,27 with the consideration of their unique advantages of high dynamic bending elasticity28, stretchability29, and high mechanic strength30,31. Despite the myriad of available working principles which the soft electronics are mainly developed from such as piezoresistive effect32,33, capacitance variation effect34, piezoelectric effect35 etc., an emerging technique triboelectric nanogenerator (TENG)36 utilizing the triboelectrification effect has enabled wide spread applications in self-powered sensors37,38. It detects mechanical stimulations from both static and dynamic processes through the voltage and current output signals of the TENG, converting ambient mechanical energy into electric energy via a coupling effect of contact electrification and electrostatic induction39,4b. The inset of Fig. 4b shows the collected corresponding current signal of foot step** (~75 nA). To quantitatively study the stretchability, strain-stress tests of SEBS, PTFE, and silicone hollow fibers were carried out by both mechanical testing machine (MTS) and dynamic mechanical analysis (DMA). As summarized in Fig. 4c, SEBS hollow fiber shows the best stretchability performance, which reaches 1900% strain. As the working distance in DMA machine was not long enough for the breaking test, strain-stress curves in Fig. 4d were used to obtain the Youngs’ modulus 1.6, 5.8, and 204.1 MPa for SEBS, silicone, and PTFE hollow fibers, respectively. Figure 4e is the output power density of a six 10-cm-conductive fiber-based TENG with different external loads under 0.1 m/s contact-separate working speed. The output current decreases with the increasing of external resistances, while the voltage follows a reverse trend (Supplementary Fig. 10), which is due to the Ohm’s law. Thus, the instantaneous power density increases firstly and then drops, being maximized at a load resistance of ∼200 MΩ, corresponding to a peak power density of ∼10.2 μW/m2. Durability is one of the key factors to ensure the long-term stability of devices, and Fig. 4f plots the current output with 2500, 5000, and 8000 contact-separation cycles under 0.3 m/s. The output keeps at around +7 nA to −10 nA, indicating a stable operation in the testing period, and the full data is shown in Supplementary Fig. 11. More durability tests of fibers with 200% strain and 1000% strain are shown in Supplementary Fig. 12 and Supplementary Fig. 13.

Fig. 4: Stretchability and durability of the super-elastic conductive fibers.
figure 4

a The liquid metal fiber is still conductive when stretched to up to 19 times. See Supplementary Movie 1. b Left: 10-cm liquid metal fibers bear the dynamic impact caused by a 1.5 kg load free-falling motion from 0.8-m height. See Supplementary Movie 2. Right: The resulting fibers can bear the impact and deformation caused by human activities such as step**. Inset: the collected current signal of foot step** (~75 nA). c Tensile test of SEBS, PTFE, and silicone hollow fibers. SEBS hollow fiber shows the best stretchability. d DMA test of SEBS, silicone, and PTFE hollow fibers to obtain the Youngs’ modulus of 1.6, 5.8, and 204.1MPa, respectively. e Power density with different out loads. Corresponding measured voltage is plotted in Supplementary Fig. 10. f Current output with 2500, 5000, and 8000 contact-separation cycles under 0.3 m/s speed to show the durability of the super-elastic conductive fibers.

Full size image

Conformal fiber net for monitoring sports performance and training

Based on the advances of its ultra-flexibility and ultra-stretchability, the fabricated conductive fiber offers an outstanding conformability and can be adapted to regular and irregular shaped surfaces. In particular, the nature of monitoring sports performance and training requires the direct contact with the training facilities,50 while maintaining stable sensing performance under sudden deformation and strong external impact. The fabricated conductive fiber is therefore a promising candidate to meet these strict requirements. So, we first constructed a fiber-based sensing net that was conformally adapted on the 2D inner surface of a baseball glove to locate the hitting points with different catching speeds. As demonstrated in Fig. 5a–c, when the ball was caught, hitting points were detected, and the catching speed was indicated by analyzing the current value differences, as lower catching speed results in a smaller current signal, and higher catching speed leads to a larger current output. (Color differs depending on the speed in the schematic in Fig. 5c).

Fig. 5: Applications as an adaptive sensor net on regular and irregular shaped surfaces for monitoring sports performance and training.
figure 5

a A fiber-based sensing net that was conformally adapted on the 2D inner surface of a baseball glove to locate the hitting points with different catching speeds. b Photo to show the ball-catching using functional baseball glove. c The detection to locate the hitting points with different catching speeds when the baseball was caught by the glove. Colors indicate different catching speeds. d Conductive fibers on a 3D surface of a football to perform the sensing function of a spherical coordinate. e Schematic to show the cross point when pressed by fingers. f–k When a cross point was pressed, the corresponding signal was visualized on the spherical coordinates. f and g Point (0°, 45°) was pressed. h and i Point (60°, 90°) was pressed. j and k Point (120°, 135°) was pressed. The color indicates the signal value, and if both coordinate values are high enough, the upright coordinate line segments or area would be highlighted, which means that this cross point is pressed.

Full size image

The second demonstration is to apply these conductive fibers on a 3D surface, such as the surface of a football, as shown in Fig. 5d, e, and this fiber net on the football surface exactly performs the sensing function of a spherical coordinate.51 Fibers attached on the football surface formed both the longitude line and the latitude line, and each cross point of the fiber was regarded as a coordinate point in the spherical coordinate system. For testing convenience, we performed the finger touch test as shown in Fig. 5e. We used the current value to detect and identify the point under impact based on the longitude lines of 0° (360°), 60°, 120°, 180°, 240°, and 300°, and latitude lines of 45°, 90°, and 135°. As the fabricated conductive fibers offered an ultra-sensitivity to environment motion change, we adopted a special data processing method to avoid possible noise jamming. When a finger touched a point, current signals from all nine lines were received. For each line, we extracted the signal values of both touched and untouched states, and then took the ratio of touched value/untouched value as the line value R. Thus, for one certain touch, we got 9 R from nine fibers. Our target was to detect which cross point was touched, so we transfered line values to the cross point values by taking the product value of the two crossed lines’ R as the point value. For instance, the detected value for point (30°, 120°) is R(lon30°, lat120°) = Rlon30° × Rlat120°. So for one certain touch, we got 18 detected values for the 18 cross points, respectively. Then, to visualize the entire sensing net, we assigned these 18 values to the corresponding positions on the spherical coordinate, as shown in Fig. 5f, g with the maxmum value appeared at (0°, 45°) and the most bright area was the upright line segment and area of point (0°, 45°). Based on this sensing principle, Fig. 5h, i show the detecting results when point (60°, 90°) was under impact, while Fig. 5j, k show the detecting results when point (120°, 135°) was under impact. Movies to show the full coordinate points on the spheres can be found in Supplementary Movie 3.

Super-elastic conductive fiber for ion movement detection in solutions

Another merit of fiber devices is that it is much easier to be packaged (only two ends) compared with film devices52,53 and on chip devices54, which is particularly suitable for underwater applications. As far as reported, TENGs working underwater mainly make use of the mechanical energy of waterwaves55, and a recently work has developed a bionic film device by collecting underwater mechanical energy through utilizing water flow in device channels56. Besides the mechanical movement induced static charge variations between/among components of TENGs, we found that ion movements and concentration changes in solutions could also result in electrical output, which is accompanying TENG waterwave mechanical energy harvesting. This phenomenon could be utilized to detect underwater electrical signals and collect related energy. Here we demonstrate a way to make use of this phenomenon based on our super-elastic conductive fibers. To build up a controllable ion movement environment in solution, we constructed a three-electrode system connected with the electrochemical workstation using a carbon cloth as working electrode, a platinum wire as the counter electrode, and Ag/AgCl as the reference electrode, as schematzed in Fig. 6f. When “+” source signal (upper figures in Fig. 6a–d) is applied, the electrochemical station gives out electrons e- with negative charges through the counter electrode (Pt wire). So, cations in the solution move from other locations toward the counter electrode, and anions distribute more at other locations in the solution. As there are negative anions outside the detecting fiber, positive charges move into fiber to reach balance. Thus, current direction is into the fiber from the electrometer, which appears as negative in the detecting signals presented in the bottom part of Fig. 6a–d. When the applied source signal is fully applied, the solution system gradually reaches a balance state, and the detected peak recovers to 0. When the “+” signal is off, an opposite process happens, and the detected signals become opposite. (Detailed discussion is shown in Supplementary Fig. 16.)

Fig. 6: Super-elastic conductive fiber for ion movement detection in solutions.
figure 6

a A 14-cm-long fiber was immersed in 0.25 M/L NaCl solution and the solution was applied with DC rectangular wave signals of 0.2, 0.4, and 0.8 V from the electrochemical workstation. The upper green line shows the source signal on-off and the below correspondingly shows the detected signal from our functional fiber. b A 10.4-cm-long fiber was immersed in 0.25 M/L NaCl solution and the solution was applied with DC rectangular signals of 0.2, 0.4, and 0.8 V. c, d A 14-cm-long fiber was immersed in 0.5 M/L NaCl solution and 0.5 M/L Na2SO4 solution under 0.2, 0.4, and 0.8 V applied DC rectangular wave signals. e Summary of a 14-cm-long fiber detecting results in different concentration solutions and different applied DC rectangular wave voltage signals. It shows that the detecting results are majorly related to the applied signal value which represents the ion movement. f A schematic of the prospected application for undersea detecting with the inset schematic to explain the self-powered detecting process.

Full size image

Factors including source amplitude, detecting length of fibers, solute types, and solution concentrations may affect the detecting results. As revealed in Fig. 6a–d, the value of sources signals obviously has an impact on the detecting results (0.8 > 0.4 > 0.2 V). So we firstly investigate how the detected signals are influenced by the immersed fiber length. In Fig. 6a, a 14-cm-long fiber was immersed in 0.25 M/L NaCl solution, and the solution was applied with DC rectangular wave signals of 0.2, 0.4, and 0.8 V through multi-potential steps mode from the electrochemical workstation. To make a comparison, in Fig. 6b, a 10.4-cm-long fiber was immersed in the same 0.25 M/L NaCl solution and the solution was also applied with DC rectangular wave signals of 0.2, 0.4, and 0.8 V. The detected signals are smaller than those in Fig. 6a under the same applied voltage, indicating that the detected signal value is majorly related to the effective fiber length immersed in solution. Next, in Fig. 6c, d, the effect of solute difference was investigated by immersing a 14-cm-long fiber in 0.5 M/L NaCl solution and 0.5 M/L Na2SO4 solution separately, under 0.2, 0.4, and 0.8 V DC rectangular wave signals. The detecting results obviously differ with different solutes under 0.8 V, and less obviously under low source signal values. Figures 6a, c investigated the solution concentration as a factor, and the detecting results did not differ obviously with different concentrations of the same solute. Full test results were extracted and summarized in Fig. 6e, verifying the above results and conclusions. Firstly, the source signal value obviously played a more important role than other factors, as the curves clearly distributed in three groups according to signal values (0.8 > 0.4 > 0.2 V). Secondly, the curves for four solutes differed little under 0.2 and 0.4 V signals, and differed larger under 0.8 V signals, but were still insufficient to be clearly distinguished with each other. Lastly, the curves did not show an obvious positive relation between the solution concentration and the detecting results. The reasons for the unclear regularity might be due to the facts that the conductivities of solutions with different solute types are different and the ion movement conditions differ with ion types (their carried charges and the ion weights may affect their movement under electric field). Therefore, this systematic study confirms that the immersed fiber length and the source signal values are the major factors affecting the detecting results. The solute type and solution concentration are the minor factors. Figure 6f demonstrates the prospected application for large-area undersea monitoring. The super-elastic conductive fibers can detect undersea electrical signals caused by ion movements and concentration changes, while collecting the energy simultaneously. The inset is a schematic to explain the self-powered detecting process.

Discussion

Ultra-stretchable SEBS hollow fibers were fabricated using the proposed soluble-core method. By incorporating liquid metal into the inner core, a super-elastic conductive fiber was achieved with easy packaging. Combining the effects of triboelectrification caused by surface contact and principles of electrostatic induction, the ultra-stretchable conductive TENG fiber can be attached to 2D and 3D surfaces working as a self-powered sensor while bearing strong and sudden impacts. Thanks to its intrinsic advantages of waterproof and easy packaging, the as-fabricated fiber was also utilized to study the electrical signal change and ion concentration variations in different solutions, indicating the potential as an underwater self-powered monitoring net. This newly proposed fabrication approach and the combination with TENG provide more possibilities to achieve various in-fiber structures and a viable path to introduce soft material enabled electronic devices into thermally drawn fibers, which allows more space for further large-area high-dimensional device integration. Especially, inspired by the effectivity of detecting ion variation related signals, multi-position underwater motion monitoring system can be expected.

Methods

Materials

SEBS pellets were G1643 from Kraton. PVA pellets were G-polymers 8049, 8042, and 8077 P from NIPPON GOHSEI which permit a temperature window (between Tm and decomposition temperature) for thermal processing. GaIn eutectic, NaCl, and KCl were from Alfa Aesar. K2SO4 and Na2SO4 were from Aladdin.

Fabrication of PVA core-SEBS shell preforms

PVA pellets were filled in the heating barrel of the injection machine, and melted at 200–225 °C for 3–5 min. Then the melted PVA was injected to the mold under pressure of 0.7 MPa. SEBS pellets were firstly hot-pressed into thin film under 190 °C for 8 mins, then cropped to rectangular shape. SEBS films were rolled outside PVA cores tightly. And the PVA cores were shaped to cylinder, cuboid, tubular, and other shapes depending on the mold shapes. For dual core preform, the two cores were wrapped separately and then combined to form one preform. The wrapped preforms were consolidated in vacuum oven at 110 °C.

Fiber drawing

The fibers were fabricated by the thermal drawing process by placing the preform in a two-zone heating furnace, where the top and bottom zones were heated to 150 and 350 °C, respectively. The preform was fed into the furnace at a rate of 3 mm min−1 and drawn at a speed of 1 m min−1, which resulted in a draw-down ratio of 18. Tens of meters of fibers were collected from each draw.

Dissolving inner core and conductive fiber fabrication

The collected fibers were cut into segments and immersed in 80 oC hot water for overnight to fully dissolve the inner cores. After completely drying off, we used a syringe to inject GaIn eutectic into the hollow channels. Then copper wires were inserted to both ends and glues were used to seal the connections.

Visualization of the sensing results on the football

To visualize the entire sensing net, a spherical coordinate system framework was constructed by 12 longitude lines × 12 latitude lines, which both supported a sphere surface shape and included all 18 detect cross points. We assigned 18 R values to the corresponding positions on the spherical coordinate, and took the detected minimum value to assign the rest coordinate points on the sphere. Then the coordinate sphere was colored depending on the coordinate point values. If both coordinate values of a point are high enough, the upright coordinate line segments or area would be highlighted, which means this cross point is pressed. The reason why the supporting lines on the sphere coordinate structure are more than we tested is that the sphere surface would be tangled when the supporting longitude and latitude lines were removed. So we just kept them and assigned them the minimum value.

Electrochemical tests

A fiber holder with some groves on the plastic plate surface (Supplementary Fig. 17) was firstly fabricated to hold the soft fiber and ajust the immersing length in solutions. The fiber was placed in the grove, and connected one end by copper wire with the working electrode from the electrochemical working station. The electrical source was applied through a three-electrode system. We used the multi-potentials test mode in the electrochemical working station to apply signals. Various concentration solutions were injected to the beaker as the electrolyte.

Characterization and measurement

A linear motor (LinMot E1100) was used to apply a periodic contact-separation working cycle. The speed was set to be 0.1 m/s. The open-circuit voltage, short-circuit charge, and short-circuit current of the fibers were measured by a Keithley 6517B electrometer, and a Stanford 570 current preamplifier. The scanning electron microscope (SEM) images were taken by JEOL field emission SEM (7600 F). The obtained cross-section optical micrographs were taken by Olympus BX51. DMA test was carried out by TA Q800. Uniaxial tensile test was examined by MTS Criterion Model 42.