Abstract
Textile fabrics have a long history of use for a wide range of applications that satisfying the needs of livelihood. In this era of modernization and high technological advancements, the use of textiles has spread to include using man-made and nonconventional textiles. Among them, hollow fiber-based textiles have great application prospects due to their high surface-area-to-volume ratio, high loading capability, high active surface area and high permeability of the porous layer. Herein, the liquid loading and release performances of various hollow fiber yarns with different parameters (i.e. spinning methods, yarn count, loading amount, releasing rate, etc.) under optimized negative pressure are presented. Results showed that the hollow fiber vortex spun yarn (with yarn count Ne 10/1) had the highest loading ability under negative pressure. Moreover, the amount of liquid loaded into the yarns first increased rapidly and then the rate became slower with an increase of loading time and finally tended to a stable value. In addition, as the release time increases, the amount of liquid released from the hollow fiber yarns became higher and the release rate became lower. These findings provide an important foundation for the development of the multi-integrated textile system, which is comprises the controllable loading and delivery of various types of active substances and chemicals.
Keywords
With the advancement of science and technology, numerous innovative textile materials have been created and one of the pioneering materials is known as “hollow fiber.” It is basically categorized into two main types, which are staple fibers and long filaments with one or more axial excavated cores. The parameters, such as the diameter, the cross-sectional shape and the number of hollows, can be modified. Generally, hollow fibers with diverse cross-sections have been manufactured by adopting customized spinnerets, such as plug-in-orifice spinnerets, segment-arc spinnerets, triple-orifice spinnerets and tube-in-orifice spinnerets. 1 The thickness of the fiber wall can also be adjusted by modifying the size of the core fluid pin of the spinneret. 2 Due to the versatile specifications and the distinctive empty core structure of hollow fibers, various desirable functions and mechanical properties have been obtained that make hollow fiber a material with endless possibilities.
Hollow fibers are widely applied in making garments, household products, materials for filtration and medical devices.3,4 In recent years, a self-care textile wearable, which was made of hollow fibers, has been developed with a drug delivery function for medical treatment. The new advantages of being simple, repeatable and potentially industrialized have been proposed.5,6 Although hollow fibers have been adopted for different aspects due to their exceptional structure of having a high surface-area-to-volume ratio, high loading capability, high active surface area and high permeability of a porous layer, more possibilities of their usage have not been completely opened up.7–11 Therefore, it is essential to unearth some other applications that are beneficial to well-being using this unique function of loading and delivery. Agriculture, which is known as one of the oldest industries, is one of the important primary resources of livelihood. It is also a crucial factor in the industrial conversion of economies.12,13 Demand for growing healthy plants is always driving the implementation and the adoption of new technologies. Recently, the application of hollow fibers has been extended to the agriculture industry. A delivery system to encapsulate and release urea in the form of layered hollow nanofibrous yarns was investigated for the slow release of plant-needed chemicals and fertilizers. 14 This has led to a new inspiration to develop a novel hollow fiber-based agrotextile system with a wide range of potential agricultural applications, such as the slow release of fertilizers, pesticides or other target liquid materials for pest prevention and crop protection. With the potential function of long-lasting release, the effectiveness of the above applications is believed to be maintained even if the amount of agrochemicals is reduced.
Conventionally, a controlled release of drugs is realized along a number of tubular hollow fibers or filaments that are contained within a membrane, for example, hollow fiber membranes for the dialysis of metabolic waste, 15 hollow fiber cartridges for pumping a cell culture and maintaining the circulation for a constant period of time,16,17 hollow fiber catheters for distributing drugs to the central nervous system and brain tissue, 18 etc. The fluid that encapsulates or passes through the lumens of the hollow fibers depends on their molecular size, the diameter of the lumens, the surface-area-to-volume ratio and the selectivity of the membranes. However, they are generally rigid in structure and lack the flexibility to be utilized in the field of soft matter engineering, such as soft robotics, flexible devices for fluid–skin interaction, surgical instruments and agricultural operations. Suits for rehabilitation is one of the examples that requires both sufficient strength support as well as not restricting natural body movement. 19 While equipping the function of drug delivery to the target site of action, textile fabrics, which are made of yarns, are the best choice. Hence, it would be promising if hollow fibers could be spun into yarns and further fabricated into different types of fabrics, so as to form a soft, light and flexible framework fitted with the curvature of the substrates. Consequently, the distance of navigation and delivery of drugs between the substrate and the fabrics can be shortened. It was reported that flexible materials are strongly preferred for making implantable devices in order to enhance the mechanical, physical and biological matching. 20 Thus, implantable textiles, which are used in conjunction with the infusion of active agents such as drugs and polymers, are potentially applied in wound healing. It is thought that the fluidic actuation could be more evenly initiated if hollow fiber-based textiles are used instead of using the traditional rigid biomaterials. In addition to medical use for human beings, agrotextiles have become widely utilized because of their excellent flexibility, durability of service life and their dimensional stability for large areas of farms. Hence, the study of hollow fibers for sustainable farm practices could be a new approach leading to effective farming, higher crop yield and better pest control, while simultaneously reducing harmful effects to the environment and the entire ecosystem. This could also contribute to advances in the agricultural transformation of farming with integrated crop protection and pest management. 22 ,23 Furthermore, the loading and the releasing of pest pheromones can be controlled by the size of lumens within the hollow fiber yarns (HFYs), in which a wide range of coverage can be secured for an even distribution of the pheromones, which conventionally rely on the effect of wind for spreading. By controlling the density of the textile, the velocity of fluid can be controlled in response to the compactness of the hollow fibers. Partial separation of fluid or gaseous substances could also be applicable.
In this study, the liquid loading and release performances of the HFYs with different parameters (i.e. spinning methods, yarn count, loading amount and release rate) have been investigated. Through the combination of both physical and natural biological controls, the findings are believed to form a strong and effective basis for the development of the multi-integrated textiles with loading and delivery functions via industrial textile manufacturing methods. Multi-disciplinary technologies, which include textile and material science, physics, chemistry and biology, are believed to play a key role in facilitating the development of the proposed system utilizing hollow fibers and textiles, which meets emerging societal needs, from the healthcare of human beings to agricultural protection (Figure 1).

Summary of hollow fiber yarns: from yarn spinning to potential applications.
Materials and experimentation
Material preparation
Staple polyethylene terephthalate (PET) hollow fibers (imported from Shanghai Different Chemical Fiber Co., Ltd) were used as the raw materials for spinning the specimens of HFYs. The yarns were spun by three respective methods, namely rotor spinning (hollow fiber rotor spun yarns (HFROYs)), ring spinning (hollow fiber ring spun yarns (HFRIYs)) and vortex spinning (hollow fiber vortex spun yarns). Deionized (DI) water and 1% (mass fraction) Maxilon Red GRL (MR-GRL) dye were utilized as the loading agents for the study of loading and release.
Liquid loading mechanism
In this investigation, the PET hollow fibers can be regarded as a tube with an extremely small radius, around 3.67 µm (statistics from scanning electron microscopy (SEM) images). Under normal (environmental) pressure, the loading capability of the samples is mainly influenced by two forces, namely the capillary force F1 and gravity force F2. As these two forces become balanced, that is, F1 = F2, the maximum height hmax can be obtained:
Method of liquid loading and release
Four experimental studies were conducted to examine the liquid loading and release capability of the three types of HFYs. All the yarn samples were produced by industrial spinning machinery, provided by Different Chemical Fiber Ltd, Shang Hai, China. Experiment 1 (E1) was conducted to compare the liquid loading performance of HFYs produced by different spinning methods, including rotor spinning, ring spinning and vortex spinning. All the yarns examined in E1 had the same yarn count, which is 20 s/1 (yarn count in the English Cotton Count system (Ne): Ne 20/1). The HFYs with the best loading performance in E1 were adopted for experiment 2 (E2). E2 investigated the relationship between yarn counts and the liquid loading capability of HFYs, and three different yarn counts (Ne 10/1, Ne 20/1 and Ne 30/1) were studied. For experiment 3 (E3), the HFYs with the highest loading capacity were used to explore the time effects during the loading process. Experiment 4 (E4) investigated the release rate of HFYs. In our study, for each parameter, we repeated our experiments 10 times to ensure the repeatability of our results.
The HFY samples were firstly pre-conditioned at 65% humidity and 21
Method for the physical examination of liquid loading and release
The DI water that contained the released colored solution was examined by an ultraviolet-visible (UV-vis) spectrophotometer to investigate the absorbance of the solution. The concentration of dyestuffs that loaded in the HFYs could then be measured. The standard absorbance was initially obtained by calibration with solutions containing specific amounts (concentrations) of dyes with the use of the UV-Vis spectrophotometer. Figure 2(a) displays the absorption spectra of the dye solutions with specific concentrations, and the absorption peak of all solutions was indicated at 506 nm. Figures 2(a) and (b) indicate that the absorbance of 62.5 mg dye per liter of water was 1.7191, where the absorbances of 31.25, 15.625, 7.8125, 3.9062 and 1.9532 mg/L dye solution were 0.8589, 0.426, 0.2175, 0.1094 and 0.0547, respectively. Moreover, as shown in Figure 2(b), the R-squared value of the above linear regression model was 0.9999, which signified the phenomenon that there was a positive linear relationship between the absorbance and concentration of the dye solution. Specifically, the higher the absorbance, the higher the concentration of dyes in the DI water. The linear equation of the model, y = 0.0275x +0.0005, was achieved correspondingly. This equation was adopted to compute the concentrations of the dye solution (the concentrations of the dissolved dyes) loaded into the HFYs in the following experiments.

(a) Ultraviolet-visible absorption spectra of dye solutions with specific concentrations and (b) A linear relationship between the absorbance and the concentration of the dye solutions was shown.
Method for studying the release rate under the swinging condition
The release rate of Ne 20/1 hollow fiber vortex spun yarns (HFVYs) loaded with dye solutions under negative pressure and the swinging condition (in order to accelerate the release process) were examined as shown in Figure 3. The Ne 20/1 HFVYs were loaded with 1% w/v solution of MR-GRL dye using the vacuum method for 1 hour. Each yarn was immersed in 200 mL DI water and subject to a swinging condition for dye liquid release. Some 1 mL of the released liquid was then extracted after the first 10 minutes and for every 20 minutes afterwards, namely at 10, 20, 40, 60, 80, 100 and 120 minutes. Then the extracted solution was diluted five times and examined by the UV-vis spectrophotometer. The characteristic absorbance peak of this examination was 506 nm. According to the result presented in Figure 3, the release rate of Ne 20/1 HFVYs rapidly increased from 0 minute to the first 10 minutes, and then slightly and steadily increased from 0.1910 (at 10 minutes) to 0.2097 (at 120 minutes), and only increased by 9.7906% of dye release from the first 10 minutes to 120 minutes. It also shows that most of the loaded solution was released from the hollows into the DI water during the first 10 minutes due to a significant pressure difference and concentration gradient between the yarns and the environment, leading to a rapid release of dye.

The ultraviolet-visible absorption spectra of Ne 20/1 hollow fiber vortex spun yarns loaded with dye solutions under negative pressure and subject to a swinging condition with different release time intervals.
Results and discussion
Morphological and structural analysis
The surface morphology of the PET hollow fibers was observed by using an optical microscope and a scanning electron microscope, as shown in Figure 4. The diameter and the length of the fiber were ∼20 and ∼30 mm, respectively. The area of the hollow was around 42.4 µm2. The hollow ratio of PET fiber is around 13.4%. These parameters ensure an effective liquid loading–releasing process with the use of the hollow fibers. In accordance with Figure 5, the counts of hollowness of fibers within a yarn were shown to be consistent by different spinning methods. It was seen that the number of hollow fibers applied in yarn spinning had not been changed a great deal even when they had undergone a series of production processes. Furthermore, the proportion of hollow areas in the cross-section of individual fibers was calculated and is shown in Figure 6. It was found that the hollow area by the method of ring spinning was the smallest, while it was the largest by applying vortex spinning. Although the same kind of hollow fibers were used, the spinning methods consist of a number of stages, such as carding, combing, drawing and twisting; thus, the hollow area of the fibers may be highly influenced by the above factors, and that may be a little varied from the original dimensions that had not been spun into yarns. Besides, the insertion of twist is one of the main differences between ring spinning (i.e. through the circulating traveler) and vortex spinning (i.e. by swirling air). The reduction of the hollow area by ring spinning might be due to this continuous process and the resulted compactness, whereas swirling air during vortex spinning might initiate a milder effect on the alteration of the hollow dimensions.

Scanning electron microscopy images of hollow fibers within the (a) vortex-Ne 20/1, (b) rotor-Ne 20/1 and (c) ring-Ne 20/1 yarns and (d) conventional polyethylene terephthalate fibers.

Hollow counts of fibers within the vortex-Ne 20/1, ring-Ne 20/1 and rotor-Ne 20/1 yarns, obtained by the method of counting the ratio of hollowness in respect to the total number of hollows in each yarn via scanning electron microscopy.

Hollowness ratio of individual fibers within the vortex-Ne 20/1, ring-Ne 20/1 and rotor-Ne 20/1 yarns, obtained by the method of approximately calculating the proportion of hollow areas in the cross-section from the individual fibers via scanning electron microscopy images.
In addition to the properties of a hollow fiber itself, in order to study the relationship between the proposed function and the yarn structure resulting from different yarn spinning methods, three kinds of yarns, which were produced via rotor spinning, ring spinning and vortex spinning, were studied. In Figure 7(a), rotor spun yarns showed a bipartite structure (two-zone structure) comprising a fiber core aligned with the helix of the inserted twist and forming the bulk of the yarn, then an outer zone of wrapper fibers, which occurred irregularly along the core length. Ring spun yarns had an ideal cylindrical helical structure uniform specific volume and each helix had the same number of turns per unit length (Figure 7(b)). Vortex spun yarns showed a structure where the strand of the input fibers was divided into two groups of relatively parallel fibers in the core and wrapping fibers at the sheath (Figure 7(c)). To summarize, ring spun yarns would possess a compact structure owing to the nearly simultaneous twisting of the parallel fibers, resulting in increased friction between them. Consequently, they are more tightly bounded together with higher fiber integration, whereas vortex spun fibers exhibited minimal compactness relatively. 24 ,25

Scanning electron microscopy images of hollow fiber yarns via (a) rotor spinning, (b) ring spinning and (c) vortex spinning.
Effects of spinning methods on liquid loading
The effects of spinning methods on liquid loading into HFYs were firstly investigated by loading 1% w/v solution of MR-GRL dye into three types of Ne 20/1 HFYs, which were produced by vortex spinning, ring spinning and rotor spinning, respectively. After 1-hour loading using the vacuum method, each yarn was immersed into 100 mL DI water and under a swinging condition for dye liquid release. The released liquid was then diluted 10 times and examined by the UV-vis spectrophotometer. According to the results shown in Figure 8, the three spinning methods shared the same absorbance peak at 506 nm. The liquid released from the HFVYs had the highest absorbance, followed by the HFROYs (medium absorbance) and the HFRIYs (lowest absorbance). As the level of absorbance had a positive linear relationship with the concentration of dye liquid, as presented in Figure 2(b), the higher the absorbance of the liquid, the more the dye contained in the released liquid. Besides, based on the equation originating from Figure 2(b), the loading masses of the HFVYs, HFROYs and HFRIYs were calculated as 9.3200, 8.2109 and 7.2437 mg, respectively, as shown in Figure 8. Since the loading mass efficiency refers to the amount of dye (mg) that can be loaded into the HFYs with a specific yarn mass, it is believed that the yarn spinning method would affect the liquid loading performance of the hollow fiber yarns, and the yarns manufactured by the vortex spinning method had the highest loading ability when compared with the rotor spun yarns and the ring spun yarns.

Ultraviolet-visible absorption spectra with the three different spinning methods.
Effects of yarn count on liquid loading
In this study, three kinds of yarns, which were produced via rotor spinning, ring spinning and vortex spinning, were studied in accordance with different basic linear densities. The yarn count is used to express how fine or coarse the yarn is. Different yarn counts (i.e. Ne 10/1, Ne 20/1 and Ne 30/1) were applied as the coarseness, bulkiness and twisting level resulting from the yarn counts have great impacts on the construction of the hollow fiber yarns, which affects the loading and release of liquid via the hollow fibers that are spun together to form the yarns.
Base on the results concluded in the Effects of spinning methods on liquid loading section, the vortex spun yarns had the highest loading ability; therefore, vortex spun yarns with different yarn counts were selected to examine the effect of the yarn count on liquid loading. The vacuum loading method is similar to the method applied above, while in this section we adopted the static releasing condition. In Figure 9(a), the absorbance peaks of the HFVYs with the yarn counts of Ne 10/1, Ne 20/1 and Ne 30/1 were found at 506 nm in the visible region. The masses of the Ne 10/1, Ne 20/1 and Ne 30/1 vortex spun yarns were 601.0, 291.0 and 194.0 mg, respectively. The liquid released (for which the liquid for loading had been diluted for 10 times) from the yarn with the Ne 10/1 yarn count had the highest absorbance (0.2581 a.u.), followed by that of the Ne 20/1-yarn (i.e. 0.2042 a.u.), while the Ne 30/1-yarn had the lowest absorbance (i.e. 0.1501 a.u.). In the meantime, the liquid concentration of Ne 10/1 HFVYs after 120 minutes of releasing was 9.3673 mg, which was 26.5% higher than that of the Ne 20/1-yarn (i.e. 7.4073 mg), while that of the Ne 30/1-yarn (i.e. 5.4400 mg) was 36.2% lower than that of the Ne 20/1-yarn. This indicated that the lower the yarn count in the English Cotton Count system (Ne), the better the loading performance of the yarns. It was thought that as a higher value in mass was defined with a lower yarn count in this indirect English Cotton Count system, the number of hollow fibers contained within a yarn was greater and more liquid could be loaded into the yarn as a result. Therefore, a negative correlation relationship between the loading ability and the yarn count was obtained. Figure 9(b) presents the release rate of Ne 10/1, Ne 20/1 and Ne 30/1 HFVYs. It shows that the rate of all the yarns was rapidly increased from 0 minute to the first 20 minutes. It also demonstrates that most of the loaded solution was released from the hollow fiber into the DI water during the first 20 minutes. Despite the similar mechanism, there was a higher releasing velocity due to a significant pressure difference and concentration gradient between the inner areas of the yarns and the environment, contributing to a rapid dye release. Notably, the surface of the yarn (fiber) is the primary source of this release. However, as time elapsed, the difference diminished, leading to a decline in the release rate. Meanwhile, the gap between hollow fibers or the interior of fibers became the major source of liquid.

(a) Ultraviolet-visible absorption spectra and (b) the release rate of hollow fiber vortex spun yarns with three different yarn counts (Ne 10/1, Ne 20/1 and Ne 30/1) and loaded with dye solutions under negative pressure.
Effects of the duration of negative pressure on loading
By comparing the results with the highest loading capacity with respect to the spinning method and the yarn count, the HFVY with Ne 10/1 yarn count (around 600 mg

(a) Ultraviolet-visible absorption spectra of loaded dye solutions and (b) the loading rate of hollow fiber vortex spun yarns for different loading periods under negative pressure.
Study of the liquid release rate
The change in UV-vis absorption spectra of Ne 10/1 hollow fiber vortex spun yarns under 6 hours of negative pressure loading and different releasing times is shown in Figure 11(a). The liquid released after 0.5 hour had the lowest absorbance (i.e. 0.2647 a.u.), while the liquid released after 2 hours had the highest absorbance (i.e. 0.2995 a.u.). For the mass release rate presented in Figure 10(b), the 2-hour release setting (i.e. 10.8727 mg) was 3.1% higher than the 1-hour release setting (i.e. 10.5455 mg). Meanwhile, the mass release rate of 1-hour setting was 9.8% higher than 0.5-hour setting (i.e. 9.6073 mg). Furthermore, the results also demonstrated that the liquid release rate of hollow fiber yarns during the first hour was faster than that during the second hour. The amount of dye solution released from the hollow fiber yarns increased along with the releasing duration and became constant and steady over a prolonged period of time.

(a) Ultraviolet-visible absorption spectra of released dye solutions and (b) comparison between the release rates of Ne 10/1 hollow fiber vortex spun yarns and those of conventional polyethylene terephthalate yarns.
When being compared with the control samples made of conventional PET yarns (fibers without hollows) (Figure 11(b)), it was discovered that the mass release rate (i.e. 16.9636 mg was already released before 0.5-hour setting) of the control samples was far higher than that of hollow fiber yarns (i.e. 9.6073 mg for the 0.5-hour setting). In addition, the released amount of the conventional PET fibers decreased after the first 1 hour, while the release amount remained constant and slowly increased for the yarn samples made of hollow fibers, despite the fact that the liquid uptake along the yarns made of conventional PET fibers is mainly via the mechanism of absorption, adhesion on the surface and capillary action, and most importantly, without “loading” into hollows. As a result, the liquid would quickly diffuse out along the concentration gradient. In contrast, a slower release could be obtained since a portion of the liquid was loaded into the lumens, in which the release dynamics was triggered or restricted by the cohesion/adhesion forces between the liquid and the fiber wall, and also under the influence of differential pressure.
Conclusion
In this article, the liquid loading and release performances from hollow fiber yarns in respect to the four aspects, namely the spinning methods, the yarn counts, the loading amount and the release rate, were investigated and verified. Through the analysis of UV-vis spectrometry, the hollow fiber vortex spun yarn with Ne 10/1 yarn count in the English Cotton Count system had the highest loading ability under negative pressure. Moreover, a positive correlation between the loading time and the amount of loaded liquid into the yarns was discovered. The longer the loading time, the greater the amount of loaded liquid. Apart from the study of loading, the amount and the rate of liquid release with regard to the releasing time were examined. It was found that the releasing amount of liquid from the hollow fiber yarns becomes higher with a longer duration of release. As for the optimization of the hollow fiber-based textile structures, the hollow fiber yarns that were produced by the vortex spinning method and with the Ne 10/1 yarn count in the English Cotton Count system were recommended to be adopted at this stage.
Although liquid can be effectively loaded and released through hollow fibers that are spun into yarns, there are some limitations that should be further investigated in the future: (i) quantifying the cross-sectional shapes and dimensions (hollowness) of hollow fibers after being twisted into yarns; (ii) quantifying the number of lumens present along a definite length of the yarns; and (iii) the variation of the fiber parameters after being processed into yarns (i.e. twisting, drawing, texturing, etc.).
It was believed that the results could form an important foundation for developing the medical applications for human beings or the inner agrochemical loaded-and-released layer of the proposed system with controllable loading and delivery functions. A wide range of agricultural textile products, such as the floating row coverings of farms and small fruit covering bags, can be hence manufactured with the application of the system. This would be an alternative to the upkeep and safeguarding of a healthy and abundant production of crops, grains, plants, vegetables and fruits from changes in the weather and attack by pests.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This work was supported by the NSFC General Program (Grant Number 32071906) and Shenzhen Science and Technology R&D Funds (Project No: 20200813100712001).
