Recent Progress on Yarn‐Based Electronics: From Material and Device Design to Multifunctional Applications

Over the last decade, the rapid development of wearable electronics has generated renewed interest in textiles. The integration of advanced nanotechnology and microelectronics with well‐established textile production processes has resulted in textile electronics, that are lightweight, flexible, breathable, and conformable, which broadens the applications of electronic products. The hierarchical textile structure, ranging from a single fiber to twisted yarns and various fabrics, is suitable for constructing multifunctional flexible devices. In particular, yarn, which bridges between fiber and fabric, is advantageous owing to its easy integration into wearable formats via weaving, knitting, or braiding. However, because of the dearth of effective interdisciplinary communication between researchers of electronic and textile engineering, fabricating yarn‐based devices with superior mechanical properties and versatile electronic functionality is difficult. Therefore, this review provides an overview of yarn‐based electronics, followed by a systematical summary of recent progress in yarns with respect to material and device design, multifunctional integration, and applications in wearable devices, including sensors, actuators, stealth, batteries, and nanogenerators. Furthermore, the major challenges and future developments in this field are discussed.


Introduction
Intelligent wearable electronics is an emerging technology formed through integrating multiple disciplines, and its engineering prototype at the laboratory level can be traced back to the concept of wearable technology proposed by professors at Figure 2. a) Schematic illustration of the twist yarns. Reproduced with permission. [39] Copyright 2015, Springer Nature. b) Illustrations of different twist processes. Reproduced with permission. [40] Copyright 2017, American Association for the Advancement of Science (AAAS). c) Schematic illustration of the hierarchically arranged helical fibers. Reproduced with permission. [41] Copyright 2019, Oxford University Press.
technologies of yarns, direct, wet, ring, air-jet, and electrospinning are demonstrated and the processes for the formation of yarns in each spinning method are discussed in detail. [36,37]

Consideration Factors for Fabricating Yarns
The conventional definition of a yarn is a fiber or filament aggregated with or without twisting, which exhibits a substantial length and a relatively small cross-section. In textile science, yarns are categorized into four types: simple, composite, blend, and fancy. Simple yarns contain only one type of fiber, composite and blend yarns contain two or more types of fibers, and fancy yarns are typically decorative with irregular structures, including slubs, spirals, chenilles, and loops appearing at regular intervals along the yarn. A yarn can be produced by twisting staple fibers or continuous filaments, in which the staple fibers are manufactured from natural staple fibers or man-made filaments that have been cut into shorter fibers. Continuous filament yarns composed of mono-or multifilaments appear smooth; however, they can be textured to induce waviness or crimps.

Twisting
Twisted architectures are ubiquitous, from molecules to nano/microscale materials and macroscopic objects such as peptides, chromosomes, and genetic materials, which are crucial in improving mechanical properties and enabling biological functions. [38] Inspired by the outstanding properties of twisted structures, a wide range of yarns with twisted structures have been reported, which can be categorized into three types: single, ply, and cable yarns. For example, two or more yarns twisted together can form a plied yarn. Introducing various multiscale gaps between helical building blocks at different levels can result in hierarchically arranged helical yarns, as shown in Figure 2a. [39] Furthermore, a combination of two or more ply yarns, cales, and ropes can result in cabled yarns, ropes, and hawsers, respectively. The fabrication of yarns involves both actual and false twisting. Any staple yarn or filament yarn must be really twisted to generate a transverse force and friction, to control the fibers inside the yarn such that they can slide under tension. Fiber slippage in yarns affects their diameter and strength, and both their diameter and strength increase with the twist density. Meanwhile, a falsely twisted yarn is formed when torque is applied to a running yarn. Although this false-twist yarn structure is not durable, it is occasionally used for specific purposes. For example, certain applications aim only at texturing yarns, whereas others aim at temporarily strengthening the strand during processing.
The twist density and helix angle are the two most important parameters for determining the characters of twisted yarn. For the former, the broadly accepted American Society of Testing Materials (ASTM) standard recommends using the units "turns/centimeter" to measure the twist density. For the latter, as numerous helix angles are involved in a yarn structure, the twist is typically defined by the helix angle of the fibers in the outermost layer. This angle is ≈45°for most typical yarns. As mentioned above, these two fundamental factors determine the yarn strength. We considered a simplified model of a staple yarn, in www.advancedsciencenews.com www.advelectronicmat.de which several fibers of a given radius exist in an approximately helical configuration. In the early stage of twisting, if none of the fibers break, then the strength of the yarn increases with a continuous twisting. However, when the twist density is excessively high, then the strength declines as twisting progresses, because the components of the fiber tension resistance are reduced as the twist angle increases, thus causing the fibers in the yarn to break. The strength-torsion relationship of the yarns can be expressed as shown in Equation 1: where y is the yarn strength, f is the fiber strength, d is the fiber diameter, Q is the fiber migration, μ is the friction coefficient, and L is the fiber length. Here, fiber migration in yarns implies that some of the outer layer fibers are entrapped in the inner layers, which results in an interlocked architecture. One of the results of helical delamination owing to fiber migration is the reduced strength of each fiber in the yarn. As shown in Figure 2b, the twisting structure can be changed by varying the twisting method, which improves the fiber strength utilization rate of the yarn. [40] Generally, twisting a bundle of fiber filaments to form a twisted yarn enhances the strength of the monofilament, thus allowing it to be woven; however, this is not necessary as some filaments possess sufficient mechanical strength, which allows them to be directly woven into the fabric. However, a low twist density is typically applied to fix the internal fibers or filaments. Generally, twisting a bundle of fiber filaments to form a twisted yarn enhances the strength of the monofilament, thus allowing it to be woven; however, this is not necessary as some filaments possess sufficient mechanical strength, which allows them to be directly woven into the fabric. However, a low twist density is typically applied to fix internal fibers or filaments. This is because an untwisted bundle of filaments can be difficult to manage owing to the projection of odd filaments and loops from the bundle surface. Nevertheless, filament yarns are occasionally twisted to a high level to degrade the luster of the yarn or to impart other attributes such as texture to the yarn. Several yarns are typically twisted together when durable and pliable yarns are necessitated, and this method is known as plying (Figure 2c). [41] Furthermore, plying in a direction opposite to that in which the component strands are originally twisted is a typical approach. If the amount of ply twisting used is sufficient to remove any residual torque, then the yarn is balanced. Such balanced plied yarns are useful for reducing postspinning processes and distortion in knit fabrics.

Poisson's Ratio
Before introducing the concept of Poisson's ratio, an introduction to mechanical metamaterials, which are a group of artificial structures with unprecedented effective properties, should be provided. [42] The classical mechanical metamaterials are generally associated with four elastic constants of the Young's modulus (E), shear modulus (G), bulk modulus (K), and Poisson's ratio ( ). Poisson's ratio is the negative ratio of the transverse strain to the axial strain and is typically a positive value ranging from 0.25 and 0.33. In nature, only a few materials have negative or large Poisson's ratios (Figure 3a). [43] However, with the development of the three-dimensional (3D) design of engineered microstructures, numerical limits are changing rapidly, and some anomalous macroscopic properties have emerged. A clear concept based on the mechanics of Poisson's ratio must be established to further develop the field of yarns.
A negative Poisson's ratio arises from the geometry of structures instead of the composition of materials. Figure 3b shows the different structures with negative Poisson's ratios. Moreover, the emergence of helical auxetic yarns with negative Poisson's ratios has birthed a new field of mechanical metamaterials. [44] As shown in Figure 3c, the helical auxetic yarn comprised a highmodulus yarn helically wound with a low-modulus yarn. Under stretching, the originally curled wrapping yarn straightened gradually, whereas the straight core yarn was squeezed and deformed into a helical shape, thus causing the overall structure to expand in the transverse direction. For example, synthetically engineered scaffolds are expected to exhibit a high negative Poisson's ratio ( Figure 3d). [45] However, this is a significant challenge owing to the necessity to satisfy the unique biological and biomechanical properties of the native tendon and ligament tissues. Natural tendons have a negative Poisson's ratio owing to the unique coiled and helical structure of collagen fibers. The formation of a lefthanded helix around the helical axis serves as a biological hinge that absorbs tensile stress and provides a spring-like mechanism to protect the collagen fibers from mechanical damage. Moreover, it presents an extremely high uniaxial tensile strength, which renders it difficult for other materials to satisfy the conditions mentioned above. [46] In summary, the auxetic effect of fiber-based materials can be attributed to fiber straightening or outward bending in response to in-plane tensile testing, which can be enhanced by weak interlayer bonding and a low network thickness.
Meanwhile, the influential concept of a giant Poisson's ratio was proposed by Pipes and Hubert, who achieved a giant Poisson's ratio by stretching an ideal helix with a twist angle of < 54.73°, in which the ratio of the fiber tensile modulus to the yarn bulk modulus was large. [47] In 2004, Baughman et al. successfully fabricated carbon nanotube (CNT) yarns with a Poisson's ratio as high as 4.2, which is important for understanding and identifying the key features of the 3D microstructures of yarns to achieve counterintuitive mechanical properties. [48] Different radial and axial distributions of the inserted twist can alter the yarn structure. For example, an inserted twist that increases the interfiber coupling can decrease the fiber slippage in the yarn and accelerate the axial elongation, which provides a basis for designing a giant Poisson's ratio when coupled with a reduction in the cross-sectional yarn area.

Shear Force
The primary purpose of determining the shear force properties of yarns is to employ the results to improve either the flexibility or twist effect. [49] The shear modulus is defined as the ratio of shear stress to shear strain, where the shear strain is generally measured in radians. [50] In one case, stress was applied to  [43] Copyright 2018, Wiley-VCH. b) A few highly simplified geometries show auxetic behavior. c) Illustration of the auxetic double-helix yarns. Reproduced with permission. [44] Copyright 2009, Elsevier. d) Schematic representation of a fibrous network. Reproduced with permission. [45] Copyright 2017, American Chemical Society. e) Schematic of the shear-band affected zone. Reproduced with permission. [54] Copyright 2018, Springer Nature. f) Scanning electron microscopy image of an adiabatic shear band. Reproduced with permission. [55] Copyright 2021, Elsevier. the diagonals of a fabric, and the shape changed according to the stress. In another case, shear stress was applied by holding one end of the fixed yarn and twisting the other end about the axis of the sample. [51] The shear modulus of a fiber can be determined via the torsional testing of the yarns. During the test, the torsional property exhibited by a yarn when subjected to a torsional force is known as torsional rigidity. Here, the torsional force is the twisting force applied to the two ends of the yarn in opposite directions. Torsional rigidity can be defined as the torque required to resist twisting, as shown in Equation (2).
where is the specific shear modulus, is the shape factor, c is the linear density, and is the fiber density.
Torsion is regarded as one of the most fundamental approaches for characterizing the mechanical behavior of fibers or yarns because it encompasses the entire deformation range, i.e., from elastic deformation and initial yielding to the strainhardening regime. Furthermore, it enables one to investigate the gradient strain effect, Hall-Petch effect, and Bauschinger effect when the sample and grain sizes are combined appropriately. [22] For example, fibers may undergo elastic, yield, strengthening, and necking stages during the tension process, and the fracture that occurs in the elastic stage is known as a brittle fracture. Brittle materials (such as ceramic fibers) typically form brittle fractures with the appearance of a fracture surface perpendicular to the tensile direction because the critical shear strength is lower than the critical tensile strength, which results in the fracture of the main chain molecules of the fibers. [52] However, the shear properties of fabrics determine whether they can be formed into complex curved surfaces. Owing to the small amount of shear deformation, nonwoven fabrics are susceptible to oblique arching when subjected to shear forces; thus, they are not suitable for making clothes.
In addition, the shear bands produced by the highly concentrated shear deformation exerting on fibers with amorphous and crystalline regions yielded different results. [53] Although yarn shear force has not been investigated directly, some related studies have been reported. For example, the room-temperature deformation of amorphous alloys is highly concentrated in nanoscale shear bands. The nanoscale shear band is accompanied by a strain field in the formation of a gradient micronscale shear band influence area, and the shear bands are superimposed through their own influence areas to elicit an interaction. Researchers concluded that shear bands are the carriers of their deformation and rheology under static loading conditions; therefore, understanding and controlling the shear bands of amorphous materials is key to elucidating the microscopic mechanism of plastic deformation (Figure 3e). [54] However, a localized plastic flow in crystalline materials under dynamic loading conditions, such as metal forming, machining, and ballistic impact, can result in the formation of adiabatic shear bands. All plastic strains are highly concentrated in the adiabatic shear band, which results in significant changes to and expansion of the structure over a short duration. By investigating the formation mechanism of the adiabatic shear band, the evolution of the thermoplastic microstructure of the material under dynamic loading at high pressures and low temperatures can be revealed (Figure 3f). [55] In addition, induced crystallization of semi-crystalline polymer materials is a typical method that utilizes the effect of the shear flow field on the crystalline region of the material.

Classification of Yarns
According to different classification standards, yarns can be divided into different types. In this section, yarns will be classified into single, fancy, blend, and composite yarn according to the composition and characteristics of yarns. This classification reflects the research status of yarns in the fiber material fields. [56] High-performance materials (MXene, graphene, and high entropy materials) face the spinning barriers. [57] The production of fancy yarns has a very high threshold because of the complexity of the adjustment of the preparation parameters. On the contrary, blended and composite spinning are popular ways that retain or improve certain properties by mixing different fibers. Especially, blend spinning has become a common method for ingeniously creating yarns by mixing high-performance materials with traditional fibers.

Single Yarns
Single yarn is made from a single kind of fiber. Compared with continuous filaments with a length of up to a kilometer, the first problem faced by centimeter-scale staple fibers is how to splice. A continuous yarn can be produced when the fibers form a firstjoint axial connection. [58] For example, Vigolo et al. prepared CNT yarns with a wet spinning technique (Figure 4a). The key point of forming the first-joint axial connection was to balance the electrostatic repulsion and van der Waals forces on the surface of the CNT wall. By adjusting the content of the surfactant sodium dodecyl sulfate in the coagulation bath, CNT fibers were aligned to form a CNT belt. Subsequently, the belt was pulled out from the coagulation bath and then twisted to obtain CNT yarns. [59] Compared with melt spinning, high-performance materials such as CNTs retain their morphology and structure through solution spinning. However, the addition of guests to form spinnable solutions is an essential step in solution spinning, which leads to the removal of the object in the subsequent process flow. Therefore, some researchers hope to obtain fiber arrays that can directly form yarns by chemical vapor deposition (CVD). In Figure 4b, Jiang et al. grew CNTs by CVD using a silicon substrate, and the super-aligned CNT arrays obtained by this method could directly pull out CNT yarns. In addition, the diameter of the yarn could be reduced by downsizing of the tip of the pick-up yarn tool. [60] The essence of super-aligned was a delicate balance between forces. When the distance between the CNT fibers in the arrays was at a nanometer constant, the phenomenon of end-to-end fiber connection occurred during the drawing process of CNT yarns.
Carbon materials have continuously brought a significant impact on the material field owing to their unique physical and chemical properties. For example, fullerenes were discovered by Curl et al. in 1985, CNTs were discovered by Iijima in 1991, and graphene was discovered by Geim et al. in 2010. [61] Currently, graphene has been used in the study of Dirac cones in topological insulators because of its special structures of energy band and first Brillouin zone. [62] However, carbon materials are difficult to be processed into other high-dimensional material forms. They are doped in the form of particles, or form a film by coating to assist other materials to improve their performance. Fortunately, the discovery of super-aligned structures not only enables the preparation of multiwalled CNT yarns, but also brings great inspiration to the preparation of other yarns, such as multifibrillar polyacrylonitrile yarn. [63]

Fancy Yarns
Fancy yarn refers to filaments that are thermally and mechanically deformed to have crimp, spiral, loop, and other characteristics. [64] Helical structure, as the most important structure, has been widely studied because of properties with extremely high stability, special topology, and compressibility. In the field of textiles, helical yarns can endow fabrics with a highly ordered structure, excellent mechanical strength, structural stability, flexibility, and biaxial stretchability.
The manufacture of helical yarns faces two major challenges. The first is that it is difficult to synthesize micro-and nano-scale yarns with high strengths, and the second is that it is hard to control the composition, length, diameter, and pitch of helical yarns. Among all the spinning techniques, microfluidic spinning technology has shown great potentials for the production of helical yarns due to its precise control of yarn structures and the continuous fabrication properties. As shown in Figure 4c, Liu et al. prepared helical yarns using microfluidic spinning technology, with formic acid, polycaprolactone, and coagulant of water as the spinning sol. [65] The as-made yarns showed spring-like stretchability and flexibility, which were much better than the reported crosslinking reaction-based micro-yarns. Furthermore, yarns could be woven into high-strength, biaxially stretched and biocompatible 3D skin scaffolds for abdominal fascial healing. On the other hand, the hydrodynamic focusing effect in microfluidics is indepth studied by Yu et al. Helical yarns were successfully prepared by forming a 3D coaxial sheath flow. During the flow of sodium alginate and CaCl 2 in opposite directions, Ca 2+ ions diffused into the Na-alginate solution and facilitated gelation of the sol. In addition, yarns with certain diameter and pitch could be obtained by adjusting the flow rate ratio, and novel helical structures including Janus, triple, core-shell, and double helical could be fabricated by changing the microfluidic injection capillary design ( Figure 4d). [66] Similar to the secondary structure, domain, and tertiary structure of protein molecular space, the multilevel spatial structures  [59] Copyright 2000, AAAS. b) CNT yarns that are continuously pulled out from a free-standing CNT array and a pulling model of yarns. Reproduced with permission. [60] Copyright 2006, Wiley-VCH. c) Actual image of the PCL helical microfibers. Reproduced with permission. [65] Copyright 2021, Wiley-VCH. d) Schematic illustration of the Janus (left) and triplex helical (right) microfibers. Reproduced with permission. [66] Copyright 2017, Wiley-VCH. e) Optical microscopy image a MXene/PU coaxial yarn. Reproduced with permission. [69] Copyright 2020, Wiley-VCH. f) Cross-section morphology of the MXene yarn. Reproduced with permission. [70] Copyright 2018, Wiley-VCH. g) Surface observations of optimal ring-spun complex yarns. Reproduced with permission. [72] Copyright 2016, Sage Publication. h) Optical microscope image of twisted NiO/ZnO yarns. Reproduced with permission. [73] Copyright 2009, AIP Publishing.
of yarns can be formed by regulating composition, orientation, and helix of yarns, thus endowing the yarns with completely new properties. The design of space structure is inseparable from Topology and Group theory. [67] These two important disciplines not only played a huge role in the molecular structure of biology for a long time, but also made great contributions to the research of condensed matter physics. Especially, Topology as a research method that abstracts objects into dots and lines, and then studies the relationship between the connection method of dots and lines and the spatial properties of objects, will definitely play an immeasurable impact on the research of yarn structures.

Blend Yarns
Blend yarn is made by mixing two or more different fibers in a certain proportion. It is a major advantage to select multicomponent fibers for the performance complementarity between materials. For example, MXene is often used as one of the components due to its excellent electronic, optical, and mechanical properties. [68] Figure 4e shows a blended yarn made by the MXene inventor Gogotsi. First, an aqueous Ti 3 C 2 T x dispersion was transferred into a mixed solution of dimethyl sulfoxide and dimethylformamide by solvent transfer method to make it compatible with polyurethane (PU). [69] After that, composite filaments were prepared by wet spinning techniques and then the filaments were collected on a roller to form a yarn after passing through a coagulation bath of isopropyl alcohol. The results showed that MXene improved the sensing performance and Young's modulus of PU fibers, and the as-made composite MXene/PU yarn could be knitted into elbow sleeves using industrialscale knitting machines to monitor several movement properties of the wearer. On the other hand, Razal et al. fabricated a flexible electrode of blend MXene/CNT yarn in Figure 4f. The MXene dispersion was drop-casted on the stacked CNT sheets. Then the sheets were twisted into yarns with a rate of ≈2000 rpm m −1 . [70] The superposition of the two high-performance materials enabled yarn-shaped supercapacitors to achieve outstanding electrochemical performance without compromising softness. In addition, its capacitance, energy density, and power density exceeded other yarn-shaped supercapacitors.
There is no doubt that material is a determined factor of yarn properties. For example, fibers are subjected to different torques due to their diverse properties during the production of blend yarns, and these changeable torques lead to varying migration lengths. Therefore, a reasonable selection of fibers can achieve uniform or nonuniform distributions such as outer tightness and inner looseness or the opposite that have totally different properties.

Composite Yarns
Composite yarn has a structure in which two or more single yarns are intertwined. [71] In many cases, it is easy to confuse composite yarns with blend yarns. The difference is that the composite yarns have a clear boundary between two isolate yarns. Usually, composite yarns can be easily obtained by twisting two different yarns. As shown in Figure 4g, Lin et al. prepared composite yarns that contained bamboo charcoal (BC), a phase change material (PCM), and stainless steel (SS) by a ring spinning frame. The composite yarns had a core-shell structure, in which the shell materials contained BC roving and PCM, and the core contained BC/SS wrap yarns. [72] Then, they fabricated elastic warpknitted fabric with excellent breathability and far-infrared/anionreleasing health care functions using the as-made complex yarns as weft yarns, rubber threads, and polyester filaments as warp yarns. The choice of the composite structure was based on the need to retain a certain function of different yarns, especially in photoelectric field. This unique structure could make the electrodes appear independent but entangled with each other. Lotus et al. successfully prepared a composite yarn with a ZnO/NiO heterojunction structure by modifying the electrospinning device ( Figure 4h). [73] The yarn had a typical rectified current-voltage characteristics and exhibited high sensitivity to UV radiation.
With the development of electronic device miniaturization technology, integrating various electronic device systems into yarns or textiles will become the mainstream in the future. The biggest problem is that different electronic devices need specific platform with certain characteristics. In order to achieve this goal, composite yarns will become an ideal candidate.

Fabrication Technologies of Yarns
In industrial production, the fabrication techniques of yarns can be divided into two categories melt spinning and solution spinning. Melt spinning is a method for converting a substance with a melting point into a viscous fluid state and then spinning, which can be divided into direct spinning and chip spinning. The solution spinning mainly includes wet spinning, dry spinning, electrospinning, microfluidic spinning, and liquid crystal spinning, the principles of which include phase separation, gel-curing, and liquid crystal formation. This section will focus on direct spinning, wet spinning, electrospinning, ring, and rotor spinning because they are widely applied in all of the fabrication techniques.

Direct Spinning
In many research papers, direct spinning and chip spinning are both vaguely referred to as melt spinning since there is not clear difference between them. Melt spinning is suitable for spinning materials that can be melted or transformed into a viscous state without significant degradation. Melt spinning is a one-component system, which only involves the heat transfer between the polymer melt sliver and the cooling medium has not any composition changes. In comparison, solution spinning is a binary (polymer and solvent) or a ternary (polymer, solvent, and precipitant) system. In the process of melt spinning, the polymer melt is extruded from the pores of the spinneret at a certain flow rate and forms filaments, which are stretched to a specific fineness between the spinneret and the winding device. [74] After sufficient cooling and solidification, the yarn is wound on a spool. Therefore, melt spinning has certain requirements for the spinneret (i.e., pore size, shape, and arrangement), cooling effect, and viscosity.
For example, Adomavičiūtė et al. prepared yarns saturated with propolis by melt spinning (Figure 5a). After melting the polypropylene polymer containing propolis in a heating zone at 210°C, the polymer sliver was extruded with a single screw extruder. Especially, the extruder had a circular spinneret with 24 holes (0.45 mm in diameter), and a speed of 20 rpm. The melt was quenched by a cross-flow air at 13°C to form filaments, which were then drawn through the drawing rollers at different temperatures and were wound up to form functional filament yarns. [75] In melt spinning, it is common to improve the spinning process by modifying the nozzle technology. Iqbal et al. successfully prepared multifilament polypropylene with different proportions of microcapsule phase change materials by using a special spinneret. To prevent damage and rupture of the microcapsule shell due to the prolonged residence time in the extrusion cavity, the holes were designed to be nonconical structure and the hole diameter was increased. In addition, the number of spinneret holes was reduced. Figure 5b shows parameters of the special spinneret. The distance between every hole was 14 mm, the spinneret  [75] Copyright 2018, Taylor & Francis. b) 2D and 3D images showing the geometry of spinneret (unit: mm). Reproduced with permission. [76] Copyright 2015, Spring Nature. c) Schematic representations of the spinning apparatus. Reproduced with permission. [79] Copyright 2007, Wiley-VCH. d) Schematic illustration of the wet-spinning process. Reproduced with permission. [80] Copyright 2020, Spring Nature. e) The fabrication of electrospun yarns by conjugate electrospinning. Reproduced with permission. [83] Copyright 2020, De Gruyter. f) A generalized twisting system. Reproduced with permission. [87] Copyright 2016, Spring Nature. g) The agent-aided system on the ring frame. Reproduced with permission. [88] Copyright 2019, Sage Publications. h) Flow diagram of disc swirl spinning process. Reproduced with permission. [90] Copyright 2009, Taylor & Francis.
depth was 12 mm, and the spinneret holes had bosses as high as 1.2 mm and apertures as depth as 0.8 mm. [76] In summary, for melt spinning, the control of the mechanical properties of yarns mainly depends on the formation of crystalline regions. Besides adjusting the cooling temperature and time, stretching rate, and other factors, introducing the extra fields such as clipping field to obtain sufficiently tough fibers for twisting has been a promising way. [77]

Wet Spinning
Wet spinning includes the following four steps: preparing a spinning dope, extruding the dope from the spinneret to form a fine stream, solidifying the dope stream into spun fibers, and directly packaging or postprocessing to form yarns. Compared to the dry spinning that uses hot air to evaporate the solvent, the wet spinning with a coagulation bath is more suitable for the production of staple fibers. [78] Moreover, the curing action of the coagulation bath enables fibers with a special skin-core structure. Therefore, wet spinning is widely used even though the spinning speed is not as fast as the dry spinning.
As shown in Figure 5c, Baughman et al. prepared a kind of single-wall CNT (SWCNT) yarns by wet spinning. [79] The homo-geneous SWCNT precursor liquid was injected into a glass tube that contained a flowing polyvinyl alcohol (PVA). The optimization of spinning parameters was achieved by adjusting the ratio of solvent and solute in the precursor solution as well as the relative speed between the spinning solution and PVA coagulation solution. The agglomerated SWCNT fibers were collected through a rotating mandrel and then drawn through a roller to form a yarn. A similar principle could also be used to prepare the MXene yarns. In Figure 5d, Eom et al. transformed 2D nanosheet MXenes into gel-fiber assemblies and realized the preparation of continuous fibers by utilizing the phase-transition ability of high-concentration colloidal dispersions in a solid coagulation bath. In particular, the spinnability of MXene dispersions was the key point to wet spinning. By increasing the content of Mxene, the dispersion viscosity increased, and the shear stress first decreased and then increased. After the ratio of the storage modulus to the loss modulus reached a specific value, MXene dispersions would be highly exfoliated, delaminated, and gelled in a coagulation bath extruded into NH 4 ions. Finally, a continuous yarn was produced by washing the spool in a water bath. [80] In summary, for both the wet and dry spinning techniques, the most critical factor for the formation of continuous yarns is to optimize the forces between fiber molecules. Compared with dry spinning, the wet spinning has advantages such as a fast production rate and can create dense structures of yarns. Nevertheless, the production rate of wet spinning is still 10 times slower in comparison with the traditional melt spinning technique for fabricating most fabrics. Therefore, how to improve the production rate of solution spinning is still a huge challenge in industrialization. [81]

Electrospinning
Electrospinning is a technology that uses the electric-field to form nanofibers. During the processes, charged polymer droplets first overcome the surface tension by electrostatic force to form a jet stream, which is then accelerated and dispersed at the top of the Taylor cone to form nanofiber precursors. Finally, these precursors are collected by the high-speed roller. It has attracted extensive attentions since it can prepare nanofibers with uniform morphology and only tens to hundreds of nanometers in diameter, which generally shows large specific surface areas. The major problem of electrospinning is that the 3D helical whipping of nanofibers during high-speed motion greatly increases the difficulty of orientating them into yarns. How to coordinately control nanofiber movement, deposition, bundling, and twisting through device optimization and spatial electric field distribution is important for forming yarns. [82] As-mentioned, many scholars have improved the electrospinning technology by developing high-speed spinning, water bath spinning, and tip-induced conjugate bundle spinning method. Considering the limited space of this review, this section will focus on the tip-induced conjugate beam method.
As shown in Figure 5e, the tip-induced conjugate beam method is a new electrospinning method developed based on the principle of tip aggregation effect, that is, the charge of the metal disk tends to be concentrated on the arc edge of the disk. [83] By applying opposite charges to the conjugate needles, the fibers stick and gather on the collector, and finally are pulled out and twisted into yarns. In detail, SiO 2 precursor solution was loaded into two syringes (with a feed rate of 18 mL min −1 ) with flat metal needles, which were connected to two opposite-polarity high-voltage supplies (±16 kV). The electrospinning process was taken at an ambient temperature of 26°C and a humidity of 56%, and the average collected yarn length was ≈12 cm. After spinning, the collected yarns with an average length of 12 cm were put into a tube furnace for calcination to obtain inorganic SiO 2 yarns. Similarly, Zhang et al. used zirconium acetate solution as a raw material, yttrium nitrate as the phase stabilizer, and polyvinylpyrrolidone as a polymer template to prepare spinning precursor solution, and zirconia yarns were then prepared by a self-assembled conjugated electrospinning device.
In addition to the above methods that need needles, a series of needleless electrospinning methods have also been developed in order to improve the efficiency of electrospinning. On the one hand, a roller-electrospinning method is proposed to achieve an effect similar to multineedle spinning. After applying the highvoltage onto the roller, several fibers are formed due to the electrostatic force of the solution in tip areas (mainly concentrated at where closest to the receiving plate). On the other hand, a nearfield electrospinning technology is put forward to simulate 3D printing. The reduction of the distance between the needle and collector will reduce the random movement of fibers, so as to control the generated filament more precisely.

Other Spinning Techniques
Traditional spinning can be roughly divided into open-end spinning and nonopen-end spinning. [84] The principle of open-end spinning is to form an open end by producing a breaking point at the continuous feeding sliver, and the true twist of sliver can be obtained by rotating the free end with the twister. The openend spinning methods include ring spinning, rotor spinning, friction spinning, eddy current spinning, electrostatic spinning, twist spinning, and pinching spinning. [85] The main difference between nonopen-end and open-end spinning is that there is no breaking point for the fed sliver, which causes the sliver without forming a true twist. Therefore, the sliver will go through false twisting, wrapping, or bonding in order to increase the mechanical character. Nonopen-end spinning includes air-jet spinning, self-twist spinning, wrapping spinning, axial spinning, and twistfree spinning. [86] This section will introduce ring spinning and air-jet spinning as classical representatives.
Ring spinning has been the most widely used spinning technology and its basic principle is described as follows. After the drafted rovings rotating into the traveler, the rotation of the bobbin causes the traveler to turn through the rovings, and then the rovings are twisted into a spun yarn. The extremely high rotational speed of ring spinning produces a huge force to make fibers intertwine inside and outside. the obtained yarns are mostly in the structure of conical helical curves, enhancing the strength. However, the high speed of rotation also produces huge friction between the bobbin and the traveler. The friction may cause the bobbin's speed smaller than the traveler speed, which leads to the problem of roving winding. In order to solve this problem and the limited productivity by twisting, many studies focused on eliminating the friction of the ring-spinning car to increase the spindle speed and reduce the twist of yarns. For example, Yin et al. changed the distribution of fiber tension and twist in the spinning triangle by introducing a false twist element between the front roller and yarn guide, thereby optimizing the structure and properties of yarns (Figure 5f). [87] A new theoretical model of yarn dynamics in a generalized twisting system was proposed to deal with two problems of twist generation and propagation. Furthermore, boundary value problems and simulation results were confirmed by the Newton-Raphson method. On the other hand, Li et al. added an auxiliary system to the ringspinning process. Because of the viscosity and surface tension of the adjuvant, the fiber ends protruding from the yarn body could be adhered to the yarn surface to reduce the hairiness of the modified ring-spun yarn, thus the tenacity of yarn was greatly improved (Figure 5g). [88] Air-jet spinning is to twist the drafted sliver by wrapping some free fiber ends around the periphery of the sliver through the jet of air. Air-jet yarn has a composite structure consisting of a core yarn with no twist (or little twist) and fibers wrapped around the outside of the core yarn. This compact structure endows the yarn with greater tensile strength because the wrapping fibers squeeze the core yarn to increase the frictional cohesion between the core fibers when yarn is stretched. Wickramasinghe et al. changed the www.advancedsciencenews.com www.advelectronicmat.de texturing nozzle of a jet texturing machine to supply steam and improve the properties of yarns. It was found that the wetting of the core spun yarn increased the tension of core yarn and the textured yarn, but decreased the loop instability, strength, tenacity, and elongation of the textured yarn. On the other hand, Gu et al. modified the original systems to improve the efficiency of air-jet spinning, as shown in Figure 5h. High productivity was guaranteed by utilizing the eddy current of the jet in the nozzle through the design of the funnel fiber conveyor, disc fiber collector, and vortex twister. In addition, the combined roller drafting system was introduced to further improve the drafting efficiency.
Making a thorough inquiry into the principles of traditional spinning has a good advantage of inspiring to current spinning technologies. For one thing, traditional spinning technologies have a mature and scalable process system including yarn feeding device, twisting device, drafting device, and winding device, et al. While new spinning technologies often only have a prototype up to now. For another thing, a series of classic parameters (such as twist point, true and false twist, twist number, draft multiple, draft rate, et al.) are not well incorporated into the new spinning technologies. Furthermore, traditional spinning technology still has great potential, which has been shown to be integrated with wearable electronics. For example, a 3D woven triboelectric nanogenerator fabric with strong shape adaptability and high elasticity is realized by 3D five-direction weaving. [89]

Fabric Manufacturing Based on Yarns
The term fabric is not restricted to products made from fibers. Unique structures made from other materials such as plastic films or metal foil can also be described as textile fabric. In this regard, this review focuses on yarn-based fabrics that are made by an interlacing process to produce woven fabric, an interloping process to produce knitted fabric, or an interweaving process to produce rope-like fabric. Three key elements of fiber, yarn, and fabric construction make integrated contribution to the performance of yarn-based fabrics. It should be noted that the air volume is a key aspect in many functional characteristics by virtue of providing a great deal of fiber mobility in yarn-based fabrics, which is required for fit and shape stability. In this part, three kinds of fabrication technology will be introduced.

Weaves
Most woven fabric consists of two sets of yarns that are interlaced at right angles to each other, and then alternately contact up and down to form a sine/cosine curve-shaped stable laminated structure. Yarns running along the length of the woven fabric are called warp ends, while those running across the woven fabric are known as weft picks. Fabric construction refers to the arrangement of these two sets of yarns, which is the most critical impact that influences the performance of woven fabrics. Plain is the most commonly used construction (Figure 6a). [91] Moreover, different performance characteristics of woven fabrics with the same construction can be achieved using many woven fabric parameters including area density, volumetric density, cover factor, and warp/weft step, in which the warp/weft step refers to the number of spaced yarns in the longitudinal (transverse) direction of the warp (weft) weave points of the same nature on two adjacent yarns. When warp and weft steps have the same count, the woven fabric is regarded as a balanced fabric. Unbalanced plain weaves commonly include satin and twill woven fabrics, which have more lustrous and smoother surface, fewer intersections and longer float, more open and softer construction than a plain fabric with the same cloth attributes by virtue of their constructions.
In addition to the basic weaves as the above discussion, other specialty woven fabrics with more complex structures have been constructed for various applications, and their geometrical texture and the design are also woven into fabrics to provide the intended performances. Examples of these constructions include dobby, Jacquard, crepe, pile, double, triaxial, and 3D woven. Specially, the principles of triaxial and 3D woven are all about introducing a third yarn on top of the warp and weft. The difference is that the three kinds of yarns in triaxial weaving are at 60°to each other in the 2D plane, and the third yarn often exists as a reinforcing yarn or functional yarn (Figure 6b). [92] While the 3D weaving is a manufacturing process in which three yarns are 90°perpendicular to each other in 3D space, and then are woven into a fabric with a certain thickness. This weaving process breaks through the shackles of flat structures and has great potential in the design of special-shaped structures.
Compared with other types of fabric, the woven fabrics have the characteristics of superior durability and high dimensional stability. Therefore, the strength and friction resistance of yarns are highly required in the weaving process. First, applying a tension to the warp yarns is a necessary step to maintain a high area density structure. Second, the lifting motion of the brown frame will cause tension fluctuations when different warp yarns are lifted in sequence. Third, the mutual entanglement of surface hairiness because of the surface friction will significantly increase the occurrence of end breaks during beating up. To solve these problems, sizing the yarn before weaving is a common method. In General, weaves have potential advantages in applications such as friction nano-generators, and its characteristics of friction resistance and high strength endow devices good washing resistance, long cycle life, and satisfactory stability. [93]

Knits
The knitted fabrics are basically fabricated by bending a continuous yarn to form a series of interlocking loops. Weft-and warpknit are two classical constructions of knits. Specifically, the difference between weft-and warp-knit fabrics is the different directions in which the intermeshing yarn traverses. Figure 6c shows the construction of the weft-knit fabrics in which each weft yarn is aligned along the fabric length. Before discussing the features of weft knits, it is important to be familiar with some basic knitting terminologies. [94] Loop is the basic unit of knit fabric, and each loop in a knit fabric is a stitch. Different stitch types such as knit stitch, purl stitch, float stitch, and tuck stitch characterize the different knit constructions. Furthermore, three knit fabric parameters (i.e., stitch density, stitch length, and gauge) are used to characterize the performance of the knit fabric with the same construction. Notably, stitch length refers to the length of yarn Figure 6. a) Structure of Plain Cloth. Reproduced with permission. [91] Copyright 2021, American Chemical Society. b) Scheme of triaxial braids. Reproduced with permission. [92] Copyright 2019, Sage Publication. c) Structure of warp knitted fabric. Reproduced with permission. [91] Copyright 2021, American Chemical Society. d) Scheme of tricot warp knit with cut piles. Reproduced with permission. [91] Copyright 2021, American Chemical Society. e) Scheme of 3D rotary braider. Reproduced with permission. [94] Copyright 2019, Elsevier. f) Actual image of the braided preform. [95] Reproduced with permission. Copyright 2021, De Gruyter.
in a knitted loop, which is a dominate factor for all knitted structures.
The four primary stitch types provide four basic weft-knitted structures including plain, rib, interlock, and purl. Plain knit is the simplest and most economical weft-knit structure with a few important features such as moderate extensibility lengthwise (10%-20%), high extensibility widthwise (30%-50%), and a potential recovery of up to 40% in width after being stretched. The rib knit is named by the number of knit and purl stitches. For example, a fabric composed of the knit and purl stitches alternated each other stitch is a 1 × 1 rib single knit. Generally, the rib knit has a higher widthwise elongation (50%-100%) and better retain warmth than the plain knit. The weft-knit structure with interlock is originally derived from rib but requires a special arrangement of needles to hide the appearance of the reverse loops, in which two sets of yarns knitting back-to-back in an alternate sequence to make the two courses of loops show Wales of face loops on each side of the fabric exactly in line with each other. Therefore, the interlock knit tends to have higher gauges, better drape and are softer when compared with the rib knit. The last type of weft knit is purl knit, which has the only structure with certain Wales containing both face and reverse meshed loops. It is achieved with double-ended latch needles or needle bed racking to transfer rib loop from one bed to the other, exhibiting the same appearance on both sides.
Knitted yarns have relatively low requirements for strength and friction resistance, but require a relatively high bending stiffness and torque resistance to endow knitted fabrics with a certain deformation ability. However, the basic warp-and weft-knit structures have a serious problem and the structural break is a chain reaction. Therefore, common methods such as introducing tuck stitch, float line structure, or transfer loops are used to support and lock the base structure to form a run-resistant fabric. The principle of the as-mentioned methods is to transfer the loops into different courses or Wales, but it has some limitations. For example, the distance of the transfer should <3 gauge because the large deformation of loop will cause the structural break. In some special cases, the distance of the transfer can be ≈5 gauge by multiple transfers, which are not encouraged since it will significantly reduce the knit efficiency. According to these reasons, a third yarn is introduced into the knit fabric and thus creates more structural possibilities (Figure 6d), such as plating stitch, laid-in stitch, and weft laid-in stitch. [95] In Generally speaking, knitted structures have great potential in flexible electronics, as knitted structure endows the textile with excellent hydrophilicity, biocompatibility, breathability, and flexibility, ensuring its wearing-comfort. [96]

Braids
Braiding refers to the process of crossing three or more yarns or chamfered strips and laying them together in the diagonal form to form narrow strips of flat or tubular fabric, including winding, threading, and knotting. Braids has a wide range of formation including medical items (chirurgical sutures), electric cables and huge ropes and tubes used in the marine oilfield sector (Figure 6f). [95] Specially, interlacing diagonally means that the yarns make an angle with the product axis in the range of 30-80°, and it is the most significant geometrical parameter of braided structures. Compared with the above two methods, the biggest disadvantage of braiding is the difficulty of mechanized production, and thus consumes manpower and time.
The principle of three yarns interlacement can be described in three steps. First, interlacing the left two yarns, where the outer left yarn goes over the next one. Second, interlacing the right two yarns, where the right yarn goes over the left next nearest yarn. Third, repeating the above steps. The principle is implemented on certain machines for braiding and the resulting product includes flat braid and tubular ropes. [89] The interlacing sequence in tubular braiding is identical to that of flat braiding, except that all the yarn ends are located around a circle and have to be kept under constant tension to produce a regular structure. This kind of sequence is similar to the maypole dance found in a number of European countries. Therefore, the classical and mainly used braiding machines are named as maypole braiding machines. The different classifications of braided machines depend on the type of the motion of the horn gears and carriers, and the track type. The machine shown in Figure 6e is a three-diagonal braiding machine. [94] Both the tracks and carriers have two opposite directions for tubular braiding. The tracks determine the path of the motion, and the carriers are rotated by the horn gears. Every horn gear is a disc with several cuts where the carrier foot fits. Once the carriers and tracks start their motions, the yarns interlace together and build the next piece of braid at the braiding point. The distance between each carrier and the braiding point does not remain constant during the braiding process and, as a result, the required length of the yarn between the carrier and the braiding point changes. The differences in the required yarn length should normally not be compensated for by yarn elonga-tion used on each carrier. Compared with as-above fabrics, the application of braids in emerging fields is rare, but its high strength, machine washability, and stability also show the potential of highimpact motion detection in flexible electronics. [97]

Yarn-Based Electronics
After the successful preparation of functional yarns, they can be processed into functional fabrics by weaving, knitting, and braiding. Compared with functional textiles formed by directly coating, spraying, and dipping on ordinary fabrics, functional textiles woven with functional yarns have the advantages of higher friction resistance, better performance, and more comfortable wearing. Although yarns and fabrics have a long history, their functions are usually limited to keeping warm, cooling, and breathable, et al. With the combination with the concept of wearable electronics, the scenarios covered by functional yarns have gradually expanded to more fields. This chapter focuses on the following five topics for wearable smart yarn and fabric electronics: actuators, sensors, energy storage, nanogenerator, and stealth.

Actuators
Unlike traditional mechanical motors, yarns/fabrics with complex structures can perform movements such as rotation, elongation, and bending through adjustable strain and stroke. Actuator yarns/fabrics can respond to changes in environmental conditions such as electricity, light, temperature, solvents, and steam, and the principle of actuation can be divided into three mechanisms of changing the materials molecular chain orders, changing the fiber volumes, and changing the yarn distance by external conditions. Molecular chain order changes are usually caused by electricity, light, and heat, and generally do not involve mass exchange and volume changes. For example, reconfiguration under thermal stimulation in shape memory polymers and alloys, electromechanical effects in dielectric materials, and phase transitions in liquid crystal elastomers. The fiber volume changes are mainly due to the absorption/desorption of small molecules caused by the potential/concentration difference, and the phase transitions such as thermal expansion, melting, or crystallization caused by the temperature difference. The yarn distance changes mainly refer to the twist shrinkage effect of the high twist yarn, the principle of which is the yarn shrinkage and rotation caused by the penetration of the object and thermal expansion caused by heating, electricity, and light. [98] In a narrow sense, the artificial muscle is closer to the intelligent actuators based on the principle of yarn distance changes. In the past decade, Prof. Baughman has continued to made great breakthroughs in this field. As shown in Figure 7a, hightwist CNT yarns were fabricated and used as the main body of artificial muscles to infiltrate filled paraffin and palladium to achieve thermal, electrical, and solution driving. [99] By coupling with different twisting methods, fast-response, large braking, and high-strength artificial muscles were successfully fabricated (Figure 7b). Subsequently, the cost of artificial muscles was reduced by simply twisting the fishing line and sewing thread. Figure 7c shows that the artificial muscle prepared by Figure 7. a) SEM micrographs of artificial muscles. Reproduced with permission. [99] Copyright 2012, AAAS. b) Static torque versus applied electrical power for yarns. Reproduced with permission. [99] Copyright 2012, AAAS. c) Applications for coiled polymer muscles. Reproduced with permission. [100] Copyright 2014, AAAS. d) Tensile stroke and nominal modulus versus temperature for a monofilament. Reproduced with permission. [100] Copyright 2014, AAAS. e) Ionic and capacitive laminate with applied voltage. Reproduced with permission. [101] Copyright 2017, Elsevier. f) CV for an ICL measured at potential scan rates from 5 to 50 mV s −1 . Reproduced with permission. [101] Copyright 2017, Elsevier. g) Optical image of a seed of pelargonium carnosum and the cross-sectional SEM image. Reproduced with permission. [102] Copyright 2018, AAAS. h) The difference between l and l p corresponding to a distance that a hygrobot advances a period. Reproduced with permission. [102] Copyright 2018, AAAS.
this method successfully lifted a 500 g of load by stroke of 12% when switched at 0.2 Hz. [100] Tensile stroke and nominal modulus versus temperature for a coiled, 300-mm-diameter monofilament muscle under 7.5 MPa static and 0.5 MPa dynamic load was shown in Figure 7d. In the follow-up work, a series of difficult problems such as energy conversion efficiency, response time, and maximum stroke were explored in depth. However, the functions of the artificial muscles are much worse than that of biological muscles. For example, biological muscles with nanoscale conversion motors driven by Adenosine triphosphate chemical energy have a high energy conversion efficiency of more than 90%, a large pulling force of ≈5 Kg per square centimeter, 30% of deformation, and a long service life. Although a series of research reports on artificial muscles far exceed biological muscles in some mechanical aspects, but the comprehensive performance is bad. In addition to seeking material innovation to achieve high-strength and large-stroke performance, twisting, as the most important part of simulating biological muscles, has an extremely significant impact on the rapid response and energy conversion efficiency of artificial muscles.
On the other hand, the principle of the fabric brake can be designed by using either a sequence change of molecular chains or a change in fiber volume. As shown in Figure 7e, Kaasik et al. developed a soft ionically conductive laminate which integrated electromechanical actuators and flexible energy storage cells. [101] By depositing an electrode layer on the fabric separator, the electrode was effectively prevented from expanding axially during the manufacturing process. At the same time, the fabric acted as a reinforce layer to prevent the short circuit problem caused by the curling of the electrode layer. Ultimately, a working voltage of no >5 V could cause positive/negative ions to shift in orientation and deflect the fabric, making it promising for use in soft robotics (Figure 7f). In addition, Beomjune et al. developed an actuatorratchet system that utilized ambient humidity energy, inspired by the bilayer structure of active plants where one layer was hygroscopic and the other was nonhygroscopic (Figure 7g). [102] The oriented polyethylene oxide nanofibers exhibited excellent hygroscopic driving effectiveness, and the rapid hygroscopic expansion of the active layer as the ambient humidity increased, deflected the actuator fabric towards the inactive layer and reached a deflection angle of 90% (Figure 7h). Furthermore, the performance of the system had been effectively improved through mathematical analysis calculations.
In summary, although textiles can respond well to environmental stimuli, the shape of yarns is more suitable from the perspective of better simulating muscles, because the muscle is formed by the cohesion of natural muscle fibers to form fiber bundles. In addition, the contraction and relaxation of muscles rely on the helical structures. [29] In the future, artificial muscles should further simulate the systolic and diastolic muscles, slowtwitch muscles, and fast-twitch muscles that exist in human muscles by twisting different yarns together.

Sensors
Integrating yarns/fabrics with conductive elements to fabricate various sensors has been regarded as a promising alternative to expensive analytical instruments used in the sports medicine or biomedical industries to monitor physiological characteristics. [103] Sensors can be divided into invasive and noninvasive ones, where invasive sensors can monitor biomarkers such as glucose, hemoglobin, and blood oxygen, and noninvasive sensors can monitor vital signals such as heart rate, pulse rate, respiration rate, body temperature, blood pressure, movement, and neuromuscular responses. Invasive sensors are subject to more stringent requirements since it needs to consider the biocompatibility of materials. Therefore, this section will select the more common and simple noninvasive sensors as the representatives.
Generally, conductive yarns for sensing can be produced by two techniques. The first one is coating sensing materials on the intrinsically nonconductive yarns, including electroplating, electroless plating, and spraying. The second one is named as twisting technique for the intrinsically conductive fibers. [104] Intrinsically conductive fibers generally refer to metal fibers and conductive polymer fibers. Although metal fibers have the highest electrical conductivity, the woven fabric is too stiff to be worn, so it needs to be used in conjunction with other fibers (Figure 8a). [105] For example, Wang et al. fabricated core-spun yarns containing CNT-coated PU and elastic cotton to detect the motion of human limbs. This yarn-based sensor could withstand up to 300% strain and could be cycled at 40% strain for nearly 300 000 cycles without fracture (Figure 8b). Although this method alleviated the problems that the metallic yarn was not easily deformed and was difficult to recover after deformation, the conductivity problem caused by the upper limit of the metallic fiber content still could not be solved. Alternatively, the technology of coating conductive materials on the nonconductive yarns can solve the problems well. As shown in Figure 8c, Cheng et al. used a plasma etching method and fabricated highly elastic PU/polyester wrapped yarns to form a corrugated texture, then coated graphene oxide on the yarn surface for reduction, and finally obtained a strain sensor with a compression spring structure. [106] The conductive yarns prepared by coating technology not only had a high electrical conductivity, but also were soft, easy to be bent, and had strong durability. Its ultra-high sensitivity to tensile strain could be used to detect a full range of human activities, from subtle to accurate speech recognition, sleep quality assessment, pulse monitoring, to human movements such as walking, jogging, and jumping ( Figure 8d).
To construct stable sensing fabrics, it requires not only sensing technology, but also a robust system design for seamless coordination between sensors, interconnects, power sources, and data storage and transmission units. [107] It is not easy to closely integrate electronic products with fabrics, simply encapsulating the system as a whole can still lead to disconnection between conductive wires and electronic devices. Therefore, the different sections of the sensing system should be protected properly according to their dimensional, mechanical, and electrical characteristics (Figure 8e). [108] For example, Tao et al. miniaturized three stretchable electrodes and screen-printed circuits, then protected them with thermoplastic PU, finally encapsulated them with polydimethylsiloxane to fabricate an electronic sensing system. The system was capable of recording high-quality signals of the electrocardiogram as well as skin temperature and respiration rate for activity monitoring, and could transmit all data to a smartphone via Bluetooth and displayed it in real-time with a specific dedicated APP. After the system was integrated into clothing, it was verified to be washable without mechanically disconnecting the electronic module (Figure 8f). The protection for the sensor is important, but it also increases the problems of signal noise and delay caused by wireless technology. How to transmit in time and accurately read the signal is a huge problem. After printing stretchable materials on flexible substrates to form capacitors, inductors, and resistors, Niu et al. solved the signal problem by introducing an unconventional RFID technology, where the wireless sensor was intentionally detuned to increase the tolerance to strain-induced changes in electronic properties (Figure 8g). [109] This technique solved the contradiction between the large strain of the resistive sensor and the small strain of the capacitive-inductive transmission signal. The design was successful in maintaining full functionality at 50% strain, allowing continuous, simultaneous, and accurate monitoring of a person's respiration, pulse, and body movement (Figure 8h).
Whether it is a resistive, capacitive, piezoelectric or inductive sensor, the first problem to be solved is the capture clarity of the signal. [28] Although there are a large number of sensors that can generate corresponding signals, they cannot fit the curved surface of the body well, resulting in a low gauge factor (GF) that makes subsequent signal transmission relatively difficult. Second, after a single strong signal is generated, other sensing Figure 8. a) Typical sensory signals for detecting knee movement. Reproduced with permission. [105] Copyright 2016, American Chemical Society. b) The morphological comparison before and after stretching. Reproduced with permission. [105] Copyright 2016, American Chemical Society. c) A schematic structure of graphene-based yarn surface after hydroiodic reduction. Reproduced with permission. [106] Copyright 2015, Wiley-VCH. d) Responsive curves of wearable sensors. Reproduced with permission. [106] Copyright 2015, Wiley-VCH. e) Schematic structure of the textronic samples. Reproduced with permission. [108] Copyright 2018, Wiley-VCH. f) Photographs of the inner sides of a textronic sample. Reproduced with permission. [108] Copyright 2018, Wiley-VCH. g) Schematic to describe a sensor node. Reproduced with permission. [109] Copyright 2019, Springer Nature. h) A body area sensor network to measure and display movement, pulse, and breathing simultaneously. Reproduced with permission. [109] Copyright 2019, Springer Nature.
indicators such as deformation degree and signal response interval need to be considered according to the application scenario. The current top-notch research has achieved fast signal response, weak deformation perception, and large working range while maintaining a relatively stable high GF throughout the entire process. Finally, its resistance to washing, friction, and bending should also be considered in order to be applied in practice.
On the other hand, the biggest challenge for fabric sensing is that as a platform for carrying the sensing system, the performance requirements of different sensing components are different. For the sensor, the fabric needs to deform enough to capture the signal. In order to achieve a high GF, the weak strain should be sensed and a stronger signal should be generated through a larger fabric deformation. While stable and weak deformation is www.advancedsciencenews.com www.advelectronicmat.de an important prerequisite to ensure that the signal is not distorted for inductive components that transmit signals. Therefore, it is one of the future development directions to rationally design yarns and fabrics with different sensing properties.

Energy Storage Devices
Incorporating flexible power sources into e-textile systems is a major challenge since there are many issues such as increased device weight and reduced flexibility of integrated textiles due to the conventional rigid battery power supply. [110] According to the principal differences, energy storage textiles can be roughly divided into supercapacitors and batteries. [27] Supercapacitors can be classified into electric double-layer capacitors, pseudocapacitors, or hybrid capacitors according to the energy storage mechanism. The double-layer capacitors are more reliable than pseudocapacitors because they can be fabricated from harmless materials including flexible carbon, metal frames, conductive polymer-coated/embedded textile electrodes, and harmless electrolytes. [111] On the other hand, the battery can be divided into chemical battery, physical battery, and biological battery according to the source, and chemical batteries can be further classified into primary batteries, rechargeable batteries, and fuel cells. Considering the convenience and eco-friendliness in practical applications, fabric-based supercapacitors, rechargeable lithium batteries, and solar cells will be introduced as examples.
Supercapacitors with much faster charging capability, higher power density, and longer cycling stability than secondary batteries, have received considerable attention in smart wearable textiles. [112] For example, according to the mechanism of surface double-layer capacitance storage charge, Qu et al. modified the proportion of hollow parts inside the fiber electrode to creat a higher specific surface area, thereby bringing a higher electrode-electrolyte contact interface and significantly promoting the charge transport (Figure 9a). [113] As a result, the hollow RGO/conducting polymer composite supercapacitor exhibited a high specific area capacitance of 304.5 mF cm −2 at a current density of 0.08 Ma cm −2 , corresponding to an ultrahigh energy density of 27.1 μWh cm −2 and a power density of 66.5 μW cm −2 , which could light the LED lamp for a long time (Figure 9b). In addition, Wu et al. synthesized uniform nitrogendoped porous graphene fibers by a microfluidic method. By increasing the specific surface area and porosity, and adding nitrogen active sites to enhance the surface charge density and interact with ions in the electrode, a supercapacitance of 1132 mF cm −2 , a high energy density of 95.7-46.9 μW cm −2 , and a high power density of 1.5-15 W cm −2 were achieved (Figure 9c). [114] When being integrated into fabrics, the yarn-based supercapacitors could successfully power various electronic devices such as audio, LEDs, monochrome displays, backlights, multicolor displays, and watches, as shown in Figure 9d.
Although fabric supercapacitors have high power densities, most of their energy densities are too low that cannot last for a long time when being used in fabric electronics. In comparison, lithium-ion batteries have the advantages of high energy density, low self-discharge, no memory effect, and wide operating temperature range. [115] However, the major problem of yarn-type batteries is the difficulty of fabricating long-length batteries, and most of the reported fibrous or yarn-type batteries only had a length of only a few centimeters. As the fiber length increased, the internal resistance became higher, which would significantly reduce the battery voltage and energy. However, He et al. confirmed that the internal resistance of fibrous or yarn-type batteries was a hyperbolic cotangent function of length, with the increase of battery length, the internal resistances were first decreased and then stabilized. Subsequently, a safe and washable fibrous lithiumion battery was successfully weaved by an industrial rapier loom (Figure 9f), which had an energy density of 85.69 Wh kg −1 and a high capacity retention of 90.5% after 500 cycles at 1 C. [116] The fabric battery could effectively charge a mobile phone or power a health management jacket that integrates a fiber sensor and textile display (Figure 9e).
On the other hand, constructing fibrous solar cells is also a plausible method to supply high-energy power for wearable electronics. solar cells can directly convert light energy into electrical energy through photovoltaic effect, which includes three processes of light absorption and exciton generation, carrier separation, and transportation of separated carriers. [117] Among the various solar cells currently available, dye-sensitized solar cells (DSSCs) have the characteristics of low production cost and high energy conversion efficiency, and the TCO-free DSSCs in the form of fabrics can widely expand their application scenarios due to their unique mechanical strength under highly deformed conditions. [118] However, in the actual weaving process, there are various problems such as excessive friction and tension, sparse fabrics, damaged electrode layer friction, limited electrode length, and difficulty in continuous fabrication. In this regard, Yun et al. proposed a new structural concept and fabrication process for DSSCs with monolithic structures (Figure 9g). [119] First, a dense fabric was wove using titanium yarns, platinum-coated titanium yarns, and fiberglass yarns as warp and weft yarns. Second, titanium dioxide slurry was printed on the fabric surface and sintered, and then the fabric was dyed and encapsulated to form electrodes. Finally, the battery was assembled by injecting electrolytes into the fabric. Of note, the DSSC performance could be optimized by tuning the fabric weave parameters, as shown in Figure 9h.
In summary, among these fibrous and fabric energy storage devices, lithium battery shows a greater potential due to its high energy density and other excellent comprehensive electrochemical performance. [120] However, these fabric lithium batteries suffer rapid capacity degradation and the formation of lithium dendrites caused by changes in the crystal structures of battery materials during the charging and discharging process, which has always puzzled many researchers. [121] For example, many commercial lithium-ion batteries use lithium-and manganese-rich cathode materials in order to further increase the battery density, which lead to rapid voltage decay and safety issues. How to improve the long-term cycling performance and safety of fibrous and fabric lithium batteries is a urgent problem to be solved.

Nanogenerators
Flexible power generator technology is used to solve the problem of physical connection between wearable electronics and power supply. According to the mechanism of power Figure 9. a) Schematic illustration of charge distributions on hollow fibers. Reproduced with permission. [113] Copyright 2016, Wiley-VCH. b) Photographs of using four connected supercapacitor yarns to power 15 LEDs. Reproduced with permission. [113] Copyright 2016, Wiley-VCH. c) The mechanism of high-performance micro-supercapacitors. Reproduced with permission. [114] Copyright 2017, Wiley-VCH. d) Scheme and photographs of the microsupercapacitors integrated into woven fabric. Reproduced with permission. [114] Copyright 2017, Wiley-VCH. e) Photographs of using a fabric lithium-ion battery to charge a mobile phone wirelessly. Reproduced with permission. [116] Copyright 2021, Springer Nature. f) Schematic structures of the fibrous lithium-ion batteries. Reproduced with permission. [116] Copyright 2021, Springer Nature. g) Photograph and schematic illustration of underlying textiles for the sample with different inter-weft spacing and number of Ti wires. Reproduced with permission. [119] Copyright 2016, Springer Nature. h) Photograph of a bent textile-based DSSCs around a rod with a 1 cm of radius curvature. Reproduced with permission. [119] Copyright 2016, Springer Nature. generation, it can be divided into ferroelectric, [122] pyroelectric, [123] piezoelectric, [124] and triboelectric. [125] Among them, piezoelectric and friction technologies have received more attention because they are easier to combine with textiles to harvest the body's heat and kinetic energy. [126] Therefore, this section will introduce the yarn and fabric-based piezoelectric and triboelectric generators as representative examples.
As one of the most common piezoelectric materials, polyvinylidene fluoride (PVDF) consists of three commonly noticed crystal phases of , and . The higher the content of -phase, the better the piezoelectric performance. To improve the piezoelectric properties of PVDF, a series of studies focus on inducing the formation of more -phase, mainly focused on the posttreatment processes such as cold drawing and polarization. However, the Figure 10. a) Optical photograph of the triaxial braided piezo generator. Reproduced with permission. [127] Copyright 2019, Royal Society of Chemistry. b) Comparison of the power density of the triaxial braided energy harvesting generator. Reproduced with permission. [127] Copyright 2019, Royal Society of Chemistry. c) Structures of the FNG. Reproduced with permission. [128] Copyright 2015, Elsevier. d) The energy harvesting of human body motion by FNG. Reproduced with permission. [128] Copyright 2015, Elsevier. e) Schematic diagram for energy-harvesting thread and SEHT. Reproduced with permission. [130] Copyright 2016, Wiley-VCH. f) Demonstration of continuously driving a smartwatch by tapping on the wrist-worn SEHT. Reproduced with permission. [130] Copyright 2016, Wiley-VCH. g) Schematic illustration of the preparation of warp and weft yarns and structure of the woven t-TENG. Reproduced with permission. [131] Copyright 2016, Wiley-VCH. h) Single-layer -TENG as self-powered wearable respiratory monitor. Reproduced with permission. [131] Copyright 2016, Wiley-VCH.
traditional manufacturing technology has many problems, such as cumbersome processing processes, slow production speed, low output power, and lack of comfort, which need to be further studied. For example, Mokhtari et al. reported a novel fabrication process for PVDF piezoelectric generators by assembling two rigid metal electrodes and piezoelectric fibers. As shown in Figure 10a, the as-spun PVDF filaments were first braided around silver-coated nylon yarns as a highly elastic inner electrode, and then the silver-coated nylon fibers were braided as outer electrodes outside the entire structure to form a triaxial structure. [127] This structure enabled the piezoelectric yarn to generate a maximum output voltage of 380 mV and a power density of 29.62 mW cm −3 , which was much better than other reported piezoelectric yarns (Figure 10b). To date, most piezoelectric yarns could be embedded in textiles for energy harvesting. Although they could not be ideally integrated into the fabric naturally and reduced the comfort, flexibility, and breathability of the fabric, they offered the possibility of a simple, low-cost, and durable piezoelectric fabric. For example, In Figure 10c, Zhang et al. reported a woven fabric, in which the composite yarns of BaTiO 3 nanowires-polymer microfibers were used as warp yarns, and electrode copper wires and insulating spacer cotton yarns were used as weft yarns. [128] After applying an electric field of 4 kV mm −1 on the interdigitated electrodes formed by copper wires for 20 h of polarization, a hybrid wearable fabric piezoelectric nanogenerator for energy harvesting was fabricated, which could generate an output voltage of 1.9 V and a current of 24 nA to power LEDs (Figure 10d).
The broadly accepted principle of triboelectric nanogenerators is contact electrification. When the interatomic distance between two materials (not limited to solid-solid) is shorter than the normal bonding length (usually 0.2 nm) in the repulsive force region, the overlapping electron clouds can cause electron transfer. [129] High-performance triboelectric nanogenerators constructed from this principle have a significant impact on the development of wearable electronics for the Internet of Things, medical science, robotics, and artificial intelligence. Professor Wang proposed a novel stretchable energy harvesting textile (SEHT), as shown in Figure 10e, and aimed to solve the problems of poor stability and nonuniformity of physically deposited metal or solution-coated conductive yarns during mechanical deformation and washing process in practical applications. [130] The SEHT, containing multistrand stainless steel wires as conductive electrodes and supersoft silicone rubber as triboelectric material, exhibited power generation outputs up to 200 V and 200 μA by stitching the triboelectric yarns on elastic textiles in a serpentine pattern (Figure 10f). Although there have been many reports on optimizing the preparation process of SEHT, the problems of low energy harvesting efficiency and poor washing resistance of fabrics still exist. Therefore, many researchers tried to improve the problems in other ways. In Figure 10g, Zhao et al. used copper-coated polyethylene terephthalate (Cu-PET) as the warp yarn and polyimide-coated Cu-PET as the weft yarn, and successfully wove a fabric triboelectric nanogenerators on an industrial sample loom. [131] A remarkable maximum short-circuits current density of 15.50 mA m −2 and an excellent washing resistance that could withstand standard machine washing tests was achieved. The SEHT could also be used as a sensor to detect signals at different breathing frequencies ( Figure. 10h). To respond to the slogans of low carbon and environmental protection, it is necessary to explore different forms of power generation models. Although the efficiencies of generating energy through these methods are low in comparison with wind power, hydropower, and solar power, the yarn and fabric generators have full potential to power the wearable electronics and will become an important supplementary energy source in the future. [26]

Stealth
Stealth yarns and fabrics can change the detectability information characteristics of targets by using various technical methods, and minimize the probability of detection by the other party detection system. According to the absorption spectrum, it can be divided into acoustic stealth, electromagnetic shielding, thermal infrared stealth, visible light stealth, ultraviolet stealth, and laser stealth. [132] This section exemplifies the shielding mechanism and research status of textiles for the most common electromagnetic radiation and thermal infrared stealth.
Based on the principle of Faraday cage, traditional electromagnetic shielding materials often use metals such as copper, steel, and aluminum to reflect electromagnetic waves, but there are problems such as heavy weight, low flexibility, high cost, and easy corrosion and oxidation. [133] Intrinsically conductive polymers such as polypyrrole, polyaniline, polythiophene, and polyacetylene could absorb waves, and show tunable electromagnetic shielding properties. Therefore, many studies have applied foam, layered, and other structures to inherently conductive polymer textiles to make electromagnetic waves reflect and absorb multiple times on the surface and inside of the material, so as to achieve the combination of conductivity and dielectric or reduce the loss of permeability (Figure 11a). [134] The reflection and absorption can be improved by using new materials, while the multiple reflections mainly involve the inner layered structure of the fabric. Similar to the preparation principle of electromagnetic shielding yarns, Zhang et al. successfully prepared a unique 3D helical yarn by in-situ polymerization of grafted polyaniline clusters after the alkali treatment of cotton yarns. The high doping level and high crystalline phase of polyaniline coating reduced the dielectric loss and microwave attenuation, and the effective improvement of multiple reflections caused by the helical structure made the electromagnetic shielding performance as high as 48.83 dB (>99.998% attenuation). Figure 11b shows the electromagnetic shielding performance of the material in the X-band. More research wove yarns into a textile to study their electromagnetic shielding performance, which could be further enhanced by the multilevel structure of the fabric. However, the preparation of electromagnetic shielding fabrics mostly used the coating technology on the fabric surface, which has a great impact on the properties of fabrics, and has the problem of easy shedding. As shown in Figure 11c, Zhao et al. developed a simple and robust synthetic route to fabricate a multilayered composite of cuproammonium fabric/polypyrrole/copper. [135] The polypyrrole coatings were first grown on cuproammonium fabrics using the dopant sodium p-toluenesulfonate and the oxidant ferric chloride, followed by electroless copper plating by chemical modification and seeding of CuO catalysts. After that, a fabric with high electromagnetic shielding performance (30.3-50.4 dB at 30-1000 MHz) was obtained, the performance of which was still excellent after standing for two years (Figure 11d). Subsequently, to further improve the electromagnetic shielding performance of fabrics in X-band, Zhao et al. fabricated lyocell fabrics with polyaniline/cobalt-nickel coatings by in-situ polymerization and electroless plating. The ternary system successfully induced the synergistic effect of electromagnetic wave absorption and electromagnetic wave reflection through the interaction of conductive polymer and metal, and reduced the magnetic and dielectric loss.
On the other hand, thermal infrared stealth refers to the technology of reducing the detectability of the target by camouflaging, reducing, and controlling the infrared band characteristic signals of the targets (e.x., the infrared bands of human body are 3-5 and 8-14 μm). The commonly used strategies include changing the infrared radiation characteristics of the target, reducing the infrared radiation intensity of the target, adjusting the propagation path of infrared radiation. [25] For example, Cui et al. developed a large-scale "freeze-spinning" fabrication technique and achieved continuous fibers with aligned porous structures by mimicking the hollow hair of polar bears. The specific steps were to extrude a well-dispersed viscous aqueous solution from a syringe at a constant speed and controlled by a programmable pump to form a stable liquid line. When the liquid thread slowly passed through the cold copper ring, the ice crystals were oriented within the thread and preferentially grew into a layered pattern, and the formed frozen fibers were finally collected by the motor and stored by freeze-drying. Figure 11e shows the changes in fiber porosity under different refrigeration conditions. [136] The textile exhibited excellent thermal insulation property, high air permeability, and great abrasion resistance, which successfully made the rabbit disappear under the infrared camera (Figure 11f). In Figure 11. a) Schematic representation of microwave attenuation by helical NaCF/PANI composite. Reproduced with permission. [134] Copyright 2019, Wiley-VCH. b) EMI shielding capabilities of NaCF/PANI, CF/PANI, and NaCF/PANI-m in X-band. Reproduced with permission. [134] Copyright 2019, Wiley-VCH. c) Schematic of the dip-coating process. [135] Reproduced with permission. Copyright 2021, Elsevier. d) SE T , SE A , and SE R of freshly prepared and 2-year-old Ti 3 C 2 T x -coated fabrics. Reproduced with permission. [135] Copyright 2021, Elsevier. e) Radial cross-sectional SEM images showing different porous structures of biomimetic fibers. Reproduced with permission. [136] Copyright 2018, Wiley-VCH. f) Optical and infrared image of a rabbit before and after wearing textile. Reproduced with permission. [136] Copyright 2018, Wiley-VCH. g) Visible-light and IR images of FIR fibers applied to clothing. Reproduced with permission. [137] Copyright 2020, Elsevier. h) Photographs that the index finger of a gloved hand wrapped in FIR fibers. Reproduced with permission. [137] Copyright 2020, Elsevier.
addition to the above-mentioned reflecting the target's own thermal infrared light for obtaining stealth effect, another effective method is emitting a specific thermal infrared light and matching the thermal infrared background around the object as much as possible to achieve camouflage. For example, Lim et al. devel-oped an active thermal camouflage textile based on far-infrared fibers (Figure 11g), which was fabricated by wet spinning using a composite solution that contained PU, tin oxide, and CNT. [137] By applying different voltages, this high atmospheric stable and long-lived yarn could realize three different infrared intensities to generate full-color infrared radiation images that could not be found in infrared cameras (Figure 11h).
In summary, fabrics have natural advantages in the field of stealth applications. By adjusting the pore sizes of the fiber-yarnfabric, it can absorb, reflect, and transmit light waves in the corresponding wavelength bands. Furthermore, these stealth fabrics can also achieve heat and humidity managements of the human body, such as thermal fabrics, cooling fabrics, and unidirectional moisture-conducting fabrics, enabling this functional fabric better wearability.

Conclusions and Perspectives
Intelligent wearable electronics is an emerging and important technology that will change the way of people's life. With the rapid development of flexible electronics, researchers begin to look for more wearable textiles as the substrate of wearable electronic devices. As the basic units of textiles, yarns, and fabrics play crucial roles due to their easy integration into wearable formats and various textile techniques. However, due to the lack of an effective interdisciplinary platform between electronic engineering and textile engineering, most of the research work focus on adding electronic functions to the traditional formed fabrics without considering the designs from the perspective of fibers or yarns. As a result, the as-made fabric electronic devices always have a lot of problems such as poor air permeability, poor durability, and insensitivity. Limited by these factors, yarn, and fabric, as important units in wearable textile electronics, have not been systematically reviewed. In this review, we systematically summarize the recent progress in functional yarns and fabrics from the perspective of yarn and fabric electronics, with special attention to the material and device designs, multifunctional integration, and their applications in wearable devices including sensor, actuator, stealth, battery, self-powering and so on.
First, the general formation principle of yarns is introduced from the perspective of twisting, Poisson's ratio and shear force. The twisting density can affect the mechanical strengths and structural characteristics of yarns, and design proper Poisson's ratio and shear force by selecting proper raw materials is crucial for constructing high-performance yarns and fabrics, since the stretch, rub and flex resistance properties are extremely important to the yarn manufacturing process. Moreover, functional materials such as luminescent or color-changing particles can be added into the yarns for achieving functional yarns. After obtaining specific yarns, fabrics with certain structures can be formed by weaving the yarns using different methods. Different textiles have different styles, properties, and structures, among which the organizational structure directly affects the stretchability, strength, air permeability, and other properties of the textiles. Through the multilevel design from single fiber to twisted yarn and woven fabric, hierarchical structures from micro to macro can be obtained, which is of great significance for the formation of multifunctional wearable electronics.
Second, the popularly used materials and yarn types are summarized in detail. As the fabrication process and application scenarios mature, the commercial yarns will slowly transition from the current single yarns to fancy yarns and blend yarns, and then to composite yarns with multifunctionality. In situ blending functional particle or powder additives materials into the fibers, yarns, and fabrics can exert their own characteristics and have the properties of softness, stretchability, and friction resistance but the homogeneous blending is hard to achieve. Coating is a useful tool to produce uniform layers of nanoparticles onto the surface of yarns/fabrics. However, it suffers from a series of problems such as an upper limit of coating cycles, limited performance improvement, and poor resistance to friction and washing.
Third, the preparation methods of yarns are systematically introduced. To sum up, solution spinning and melt spinning are the most two popular methods for fabricating yarns. Compared with the former, the melt spinning has special requirements for the raw materials, which must be melted but cannot be decomposed at a certain temperature for spinning. Nevertheless, the technological process of melt spinning is much simpler, the spinning speed is faster, and its scalability is higher than that of solution spinning in terms of industrialization potential for fabricating yarns. On the other hand, the highly mature traditional spinning methods have also made surprising progresses in certain yarn-fabrication fields. In the future, it is expected that the breakthrough of new spinning technology can be better promoted through a deeper understanding of the typical two representative principles in the traditional spinning methods.
Finally, interesting and important applications of yarns and fabrics in actuators, sensing, energy storage, power generation, and stealth are reviewed. It can be concluded that the potential of developing yarn and fabric-based electronic devices remains to be explored, and the lack of deeply study on yarns has led to many problems in terms of durability and multifunctionality. The critical application problems in these devices can be summarized as low strength and stretchability, lack of human-friendly design as well as the weak friction and washing resistance.
In addition to the above three major points that need to be considered for constructing high strength yarns and fabrics, the reinforcement for nano-scale nonwovens is also a popular research direction since these nano-scale nonwovens are widely used in biomedical, battery energy storage, sensing, and detection due to their special electrical, thermal, and magnetic properties, as well as large specific surface areas and porosity. However, the random arrangement of fibers and the weak interaction between fibers lead to great room for improving strength, stretch, and resilience of nano-scale nonwovens. There have been many studies on the reinforcement and toughening of these nano-scale nonwovens by adjusting the raw materials and farbic structures. For example, the flexibility, bending stiffness, and tensile strength of the film can be improved by in-situ cross-linking reactions between different molecules, by changing the structural, or by directly covering the surface with a protective layer of nanoparticles. However, the above methods have limited improvements on enhancing the strength from the level of MPa to GPa. Up to now, many studies have made great progress in the fabrication of single-fiber devices, yarn devices, and fabric devices for wearable electronics. In addition, it has been confirmed that different twisting structures have great effects on the yarn properties, and the tissue structure of fabric will also affect its piercing, stretching, bursting, and other properties. Therefore, it is foreseeable that with the further integration of the textile field and the electronics field in the future, there will be more reports on the impact of yarn and fabric structures on wearable electronic devices.
On the other hand, in the previous research on smart yarns and textiles, many significant innovative research focused on new materials, preparation methods, the fancy applications, and industrialization. First of all, the foundation of fabric is yarn, and the foundation of yarn is fiber. It is one of the main challenges to choose the appropriate principle to realize the special structure and high performance according to the characteristics of fiber materials. For example, the multistage plasticization principle can be used to eliminate the wrinkles on the graphene surface to achieve orderly and high-density stacking, the performance complementarity can be achieved by forming materials with different properties into a core-shell structure, and high-entropy ceramic fibers can be formed by mixing various elements. Second, it is also a challenge to solve the problems encountered in actual production by optimizing the preparation process. Many scalable methods that have not been mentioned above also show potential for fabricating wearable electronic textiles. For example, A needle-free vortex solution blowing system based on the principle of Karman Vortex Street, a sol-gel method to form a 3D network structure by hydrolyzing and condensing liquid-phase precursors containing highly active components, and microfluidics technology for precise control and manipulation of microscale fluids (especially submicron structures). Microfluidics, also known as lab-on-a-chip, is widely integrated with many technologies because of its scalability. For example, with the microfluidic blow-spinning technology, researchers have successfully produced ligand-free perovskite quantum dots on a large scale. At present, the electro-microfluidic spinning technology has been successfully applied in the field of quantum dot preparation, and fibers with a porous sponge outer layer and a dense aerogel inner layer can be prepared by combining the electro-microfluidic spinning with wet spinning. Besides microfluidics, there are other combinations between fabrication techniques, such as melt electrospinning and dry-wet spinning. Therefore, the integration of preparation methods is also one of the future development directions. Thirdly, great progress has been made in the development of multifunctional integrated wearable electronic devices, such as the triboelectric nanogenerator fabrics with sensing capability. In the future, wearable electronic textiles will integrate various functions.
As society moves towards intelligence, people's interest in wearable technology will continue to expand. Industrial production is an important part that cannot be ignored before it can be put into practical application. How to improve the production line to achieve high efficiency, low cost, and stable performance of products is a problem faced by industrialization. In addition to the need for in-depth understanding of fibers, yarns, fabrics, and electronic products, modeling technology will be gradually integrated into the textile manufacturing process. By parameterizing the structural characteristics of the textile process, it is helpful for decision-makers to find the best solution to understand the complex relationship between various textile process parameters and performance characteristics. To sum up, although the development of intelligent wearable textiles and functional yarns still faces great challenges, we believe that with the continuous efforts of researchers in different fields and the integration of multiple disciplines, the insurmountable problems will be solved soon.