A Facile and Strategic Approach to Superhydrophobic Fibrous Structure with Biaxially Aligned Electrospun Porous Fibers

Electrospun fibrous structures can be developed superhydrophobic while remaining breathable. However, current electrospinning‐based methods for developing superhydrophobicity require complex fabrication processes and multiple raw materials, lacking a facile approach to superhydrophobicity using electrospun fibers. Inspired by a cultural relic known as Plain Gauze Gown, a fibrous structure consisting two layers of porous fibers aligned in warp and weft directions by simple electrospinning is developed. Through investigation of wetting behavior, an unique Cassie–Baxter‐“restoring” (CaRe) wetting contributing to a stable Cassie–Baxter state is unveiled in the developed structure. The CaRe wetting ensures that droplet stays on the upper‐layer fibers even at a sparse inter‐fiber distance, enabling strategically reaching superhydrophobicity by lowering structure solidity. Remarkably, the developed structure with water contact angle between 159° and 162° and roll‐off angle from 10° to 3° has a water vapor transmission rate of 20.8 kg m−2 d−1, which is the highest value among all superhydrophobic electrospun structures reported, and it is also waterproof and semi‐transparent. These features make the structure suitable for a wide range of applications, including developing waterproof, breathable, and superhydrophobic membrane with simple preparation and low cost, and as a surface layer for wearable electronics that facilitates sweat evaporation and prevents water intrusion.


Introduction
In ancient Chinese literature, a special fabric known as "Jiaoxiao Gauze" was described as being "light as a feather, into water it goes, but dryness it retains".Although this legendary fabric has never DOI: 10.1002/admi.202300507been seen, the literature expresses people's genuine desire for a textile that is both featherlight and hydrophobic, offering breathability for hot weather and resistance against wetting by raindrops.Even today, such fibrous materials are still highly sought after for various modern applications that require a breathable and superhydrophobic surface property, including protective clothing, wearable electronics, wound dressing, and gas sensors. [1]ortunately, the production of fibrous materials akin to the long-awaited "Jiaoxiao Gauze" becomes possible nowadays through the utilization of a cutting-edge textile technology known as electrospinning, which is a facile and versatile technology to fabricate microand nano-fibers. [2]By controlling the fiber material and surface morphology, the resulting fibrous structure can exhibit sufficient surface roughness and low surface energy, enabling it to achieve superhydrophobicity (which is characterized by a water droplet contact angle >150°and a roll-off angle <10°) [3] while remaining breathable.
However, current electrospinning-based methods for developing superhydrophobicity usually require complex fabrication processes and multiple raw materials, which increase the overall cost and hinder them from broad application.For example, synthesizing and applying fluorinated polymers into electrospinning solution, [4] adding nanosized additives to fiber surface, [4b-d] and various post-treatments to incorporate a secondary material, such as poly(perfluoroalkyl ethyl methacrylate) [5] and dichlorodimethyl silane [6] with desired surface energy (see Table S3, Supporting Information) are commonly reported in the literature.
Moreover, a strategic approach to superhydrophobicity in electrospun fibers is still lacking.Current theoretical works have limited abilities to guide the development of superhydrophobic electrospun fibrous structures because either the works are limited to modeling the stability of wetting state or the theoretical structures are difficult in replicating.For example, Rawal [7] reported an analytical model relating fiber parameters with the pressure difference across liquid-vapor interface.The influence of fiber diameter and inter-fiber distance on the Cassie-Baxter (CB) wetting stability is quantified.Emami et al. [8] simplifies two-layer fibrous structure as a plane to explore the influence of fiber orientation on the CB stability.In addition, Onda [9] built a model for the equilibrium contact angle on fibrous structure, which is simplified as a porous film, so the equilibrium contact angle is related to volume porosity, film thickness, etc.
As a result, the control of fiber morphology parameters, including fiber diameter, inter-fiber distance, and fiber orientation, in previous works has often relied on trial-and-error approaches.As a result, the long-awaited realization of a "Jiaoxiao Gauze"-alike material has still not yet been achieved because a fibrous structure often thicker than 50 μm [10] is needed to reduce the heterogeneity of fiber distribution.
The objective of this study is to develop a superhydrophobic electrospun fibrous structure to resemble "Jiaoxiao Gauze" by a strategic approach based on the understanding of the wetting behavior on fibrous structures.The structure development is inspired by the closest culture relic of "Jiaoxiao Gauze" known as the Plain Gauze Gown (discovered in 1972 from the Mawangdui Tombs of the Han Dynasty (206 BC-9 AD) in Changsha, China).We identify the Plain Gauze Gown as a solid foundation to develop fibrous structures with superior hydrophobicity.Specifically, its low solidity resulting from the sparsely arranged fibers (see the microscope image in Figure 1a) reduces the contact area between droplets and the fibrous surface, leading to a spherical shape of droplet with a large contact angle ( c ).Moreover, the directionally aligned fibers could serve as a track for droplet to roll off the surface at a small tilting angle ( r ).
Herein, we report an electrospun fibrous structure consisting of two layers of porous polystyrene (PS) fibers aligned in warp (i.e., longitudinal) and weft (i.e., transverse) directions, denoted as biaxial structure.The biaxial structure unveils a unique "Cassie-Baxter-restoring" (CaRe) wetting, allowing for strategically approaching superhydrophobicity by lowering the solidity of structure.Remarkably, the resulting fibrous structure exhibits superhydrophobicity with  c between 159.3 ± 0.9°and 161.5 ± 0.7°,  r from 10 ± 1°to 3 ± 0.6°, and superior breathability with a water vapor transmission rate of 20.8 ± 0.1 kg m −2 d −1 .Potential applications are demonstrated at the end of this study for the development of waterproof breathable and superhydrophobic membrane, and the surface layer for protecting wearable electronics.

Wetting Behavior
Figure 1a,b shows the electrospinning setup and the morphology of the produced fibers, respectively.As shown in Figure 1a, a rotating fin collector was used to align the electrospun fibers.The biaxial structure was produced by rotating the substrate on the collector after depositing lower-layer fibers.The alignment of fibers was driven by the mechanical stretching exerted by the rotating collector and the electrostatic force between adjacent fins. [11]In addition, aligned fibers can also be produced by gap electrospinning. [12]Moreover, pores on fiber surface were created by electrospinning at controlled relative humidity (RH) based on the breath-figure: [13] water vapor was condensed on the fibersurface and created pores before the fiber was fully solidified.The SEM images in Figure 1b show the morphology of electrospun fibers at RH levels of 40%, 50%, 60%, and 80%.The average pore size and the fraction of the fiber-surface area occupied by pores increased from 77 ± 16 to 180 ± 49 nm and 34 to 51%, respectively, with RH from 40% to 80%.Then the optical microscope images in Figure 1b show that the electrospun fibers are arranged in two different alignments: uniaxially and biaxially aligned.The uniaxial alignment has been commonly adopted in previous works without controlling inter-fiber distance, [14] whereas the biaxial alignment is developed in this work to mimic the structure of the Plain Gauze Gown.
The produced porous PS fibers were deposited on a substrate surface to study the wetting behaviors in terms of apparent contact angle and roll-off angle along the upper-layer fiber direction (see Figure 1c).It is worth to know that anisotropic wetting behavior exists because of the directional alignment of fibers (see Section S1, Supporting Information for details).Before systematic data collection, a preliminary test confirmed that the pores on fiber surface can improve hydrophobicity (see Section S2, Supporting Information).This result concurs with earlier works (e.g., [5a,15] ) aiming at enhancing hydrophobicity by introducing porous structures.Thus, all fibers are porous in the following studies.
To lay the basis for a strategic approach to achieve superhydrophobicity, the wetting behavior on the surfaces dressed by uniaxially aligned porous fibers were studied first for the simplicity of structure.The grey scatters in Figure 2a show the measured contact angle and roll-off angle of water droplets on uniaxial structure with fiber diameter of 3.95 ± 0.25 μm.The results show that increasing the average inter-fiber distances (l) from ≈9 to 18.2 μm improves hydrophobicity, as evidenced by the increased  c and the decreased  r from ≈145°and 55°to 150°and 35°, respectively.The increase in  c with enlarging l indicates a Cassie-Baxter (CB) state because this finding can be well explained by the CB model. [16]Specifically, in the CB state, the droplet is supported by fibers without touching the underlying substrate.As a result, increasing l in CB state decreases the liquid-solid contact fraction, f, thereby increasing the  c (see Equation S2.2, Supporting Information).This decrease in f also explains the observed decrease in  r because decreasing f reduces the friction that impedes droplet from rolling.Therefore, the droplets were observed to roll off the testing surfaces at smaller  r with increasing l.In the rest of this study, this correlation between  c ,  r , and f is used to interpret the change in hydrophobicity for fiber-dressed surfaces before the liquid fully wets the substrate (i.e., hydrophobicity improves with decreasing f and vice versa).
Further increasing l from 18.2 to 33.5 μm weakens hydrophobicity, as evidenced by the decrease of  c from 154.5 ± 0.4°to 130 ± 0°and the increase of  r from 32 ± 0.6°to a sticky state even at 90°tilting (see the corresponding insets in Figure 2a).This result indicates that a wetting transition occurs at l >18.2 μm for the surfaces dressed by uniaxially aligned porous fibers: at this l, liquid starts filling up some of many inter-fiber grooves under the droplet; and further increasing l leads to more grooves filled by liquid and eventually results in the Wenzel state where the droplet fully wets the underlying substrate.The wetting transition weakens hydrophobicity because the liquid-solid contact fraction f greatly increases once the droplet touches the underlying substrate, enhancing the adhesion between the droplet and the covered area.Therefore, superhydrophobicity is not attained by the uniaxial structures because of the wetting transition.
Then the wetting behaviors on surfaces dressed by biaxially aligned porous fibers with fiber diameter of 3.95 ± 0.25 μm were studied, and the measured contact angle and roll-off angle were presented as the blue scatters in Figure 2a.The wetting behavior on the biaxial structure deviates from that on the uniaxial structure when the average inter-fiber distance (l) exceeds 18.2 μm.On the contrary, wetting transition is not observed for droplets on the biaxial structure for l between 18.2 and 34 μm: the hydrophobicity continues to improve without a sudden change for all the l tested (see corresponding insets in Figure 2a).
The experimental results are unexpected for the biaxial structure in Figure 2a.Note that the biaxial structure is essentially a stacking of two layers of uniaxially aligned fibers in the warp and weft directions, and thus the droplets on the biaxial structure are expected to behave in the same way as on the uniaxial structure before reaching the lower-layer fibers.One would then intuitively expect that the wetting of the biaxial structure would transit to a state where the droplet contacts the lower-layer fibers once the average inter-fiber distance (l) exceeds 18.2 μm.Because the same droplet on the uniaxial structure would have touched the substrate.Had the droplet touched lower-layer fibers, the measured  c and  r should have suddenly dropped and risen, respectively, because of the increase in liquid-solid contact fraction f caused by the contact between the droplet and lower-layer fibers.This weakening of hydrophobicity is expected to occur, even if it may not be as severe as seen in the uniaxial structure when l >18.2 μm.However, the hydrophobicity of the biaxial structure maintains an unchanged trend for the tested l range of 8.7 to 34 μm.This unchanged relationship suggests that the droplets did not touch the lower-layer fibers, i.e., the CB wetting state is maintained.
In addition, evidence from the free evaporation of an identical droplet on the biaxial structure further confirmed the preced-ing argument that the droplet is in the CB state for l > 18.2 μm without touching the lower-layer fibers.As shown in Figure 2b and Video S1 (Supporting Information), the capillary-induced deformation of liquid edges and the sliding of the two ends of the meniscus along the upper-layer fibers indicated that the droplet was in contact with upper-layer fibers (see the yellow dot-lines).The interaction between the droplet and the lower-layer fibers was not observed until the last few seconds of the evaporation, when one end of the meniscus slid along the lower-layer fibers (see the red dot-lines).The video clip (Video S1, Supporting Information) more clearly showed that only the upper-layer fibers were in contact with the droplet until the last seconds, which was evidenced by the noticeable pulling of the upper-layer fibers by the receding liquid edges.Therefore, the evaporating droplet was in CB state until the last few seconds.Note that this CB state is more stable prior to evaporation because the evaporation is essentially a trigger that weakens the stability of CB state and increases the tendency of wetting transition to Wenzel (W) state (i.e., evaporation-triggered wetting transition). [17]In summary, Figure 2a,b and Video S1 (Supporting Information) strongly evidence that the droplets on the biaxially aligned fibers were in the CB state without touching the lower-layer fibers.
To better understand the mechanisms behind the preceding unexpected experiment phenomenon, we mapped our sampling points into the wetting state chart of the uniaxial structure.By identifying the location of the observed wetting transition for the uniaxial structure on the chart, we can determine the cause of this transition and gain insights for further investigating the biaxial structure.Figure 2c shows the wetting states of a 3 μL droplet on the uniaxial structure with different inter-fiber distances (l).The CB region, i.e., blue region in Figure 2c, was obtained by theoretical derivation (see Section S3, Supporting Information), and the result shows that the CB state dominates for l < 8.7 μm because the CB state is thermodynamically more stable than the W state.In contrast, as simulated by the numerical model, l greater than 107.4 μm (see the dark grey region in Figure 2c) results in W state because the bottom of the liquid-vapor interface reaches the level of the fiber bottom (see the inset in Figure 2c).Thus, wetting transition from CB to W state due to sagging effect [18] is expected when in contact with the substrate.However, there a metastable region where the CB state could sustain despite being thermodynamically less stable than the W state [19] (see Section S3, Supporting Information).The metastable CB state occurs if the energy acquired by the droplet from environmental disturbance, such as vibration, air diffusion, and impact, cannot overcome the free energy barrier between the CB and W states [20] (that arises from the re-entrant structure of fibers and the hydrophobicity of the material). [21]As shown in Figure 2c, the experimentally observed wetting transition for the uniaxial structure falls into the metastable region, which indicates that the energy barrier separating the CB and W states of the uniaxial structure was overcome by the environmental disturbances once l is >18.2 μm.Based on this analysis, it can be inferred that the advantage of biaxial structure over the uniaxial structure in resisting wetting transition, as observed in Figure 2a, can be attributed to its greater energy barrier separating the CB and W states compared to the uniaxial structure.However, this argument still leaves a question of why the droplet is in CB state on the biaxial structure instead of a transitional state between CB and W that touches the lower-layer fibers.To answer this question, we further mapped a more detailed energy profile for droplets wetting surfaces dressed by fibrous structures in the following section.

Energy Profile
The energy profile for a water droplet wetting a fiber-dressed surface was obtained by numerically solving Equation (1) [22] in the developed model: where E is the total interfacial energy; A and  are the surface area and interfacial tension, respectively; subscripts L, V, F, and S represent liquid, vapor, fiber, and substrate, respectively;  Y,F and  Y,S are the intrinsic contact angles of the porous fiber and the substrate materials, respectively.A stable or metastable wetting state is identified when E is at a global or local minimum.The droplet volume, fiber diameter, and pore area fraction on the fiber-surface in the model were 3 μL, 4 μm, and 0.5, respectively, which were consistent with those in the preceding experiment (Figure 2a); the inter-fiber distance (l) was set as 30 μm because the experimentally observed wetting behaviors between the biaxial and uniaxial structures were notably different for l > 29 μm (see the dot-line in Figure 2a, and the insets showing droplets pinning on the uniaxial structures while rolling on the biaxial structure).Based on the developed numerical model, we particularly studied the free energy of droplets in CB, W, and additional two transitional states to map the energy profile for droplet wetting fibrous structures.As illustrated in Figure 3a.State(i) is the transitional state in which the droplet just starts to encounter the solids beneath the upper-layer fibers.The solid is the substrate surface for uniaxial structure or the lower-layer fibers for biaxial structure.State(ii) represents any transitional states between State(i) and W state, e.g., States(iiʹ) and (ii″), in which the liquid partially wets the beneath solids.The free energy of the water droplet at CB, W, and two transitional states is denoted as E CB , E W , E i , and E ii , respectively.
The grey dash-dot-line in Figure 3b illustrates the mapped energy profile derived for wetting surfaces dressed by uniaxially aligned porous fibers.The profile was mapped as E i > E CB > E W because the modeling results identified CB and W as the wetting states with the lowest free energies before and after the droplet touching the substrate, respectively, and the normalized energy difference between CB and W states was 15176.4.In addition, the free energy of State(i) was at the peak because State(i) is a transitional state with the largest liquid-vapor interface (see Figure 3a and Equation 1).As a result, the droplet transits to W state on the uniaxial structure with an average inter-fiber distance of 30 μm (see Video S2, Supporting Information) after overcoming the free energy barrier because E W is the lowest among all wetting states, which concurs with the experiment results in Figure 2a.
However, the mapped energy profile for wetting the biaxial structure (see the blue and orange lines in Figure 3b) was different than wetting the uniaxial structure (grey line).The profile was mapped as E i > E iiʹ > E CB by studying two consecutive wetting sub-processes, in which droplet wetted only the upperlayer fibers (i.e., sub-process 1), and both upper and lower-layer fibers (i.e., sub-process 2).The sub-process 1 is the same as the process wetting the uniaxial structure before liquid reaches substrate, thereby, we can map E i > E CB .For the sub-process 2, the modeling results show that there is a metastable state, i.e., State (iiʹ), soon after State(i) with a local minimum E iiʹ , thereby, E i > E iiʹ .In addition, the obtained E iiʹ is greater than E CB because ∆E iiʹ−CB = 9.7.Furthermore, the energy required for the droplet to continue wetting the biaxial structure beyond State(iiʹ) is greater than the energy peak at State(i).As a piece of evidence for this argument, we identified a follow-up state, denoted as State(iiʹʹ), with a higher free energy than State(i); and the droplet has not yet reached the underlying substrate at State(iiʹʹ) (see Section S4, Supporting Information).This finding is consistent with the preceding analysis for Figure 2c that the overall energy barrier between CB and W for the biaxial structure is greater than that for the uniaxial structure.
The derived energy profile in Figure 3b for wetting the surface dressed by biaxially aligned porous fibers explains the unexpected phenomenon in Figure 2a for the average inter-fiber distance between 18.2 and 34 μm, that is even if the droplet touches the lower-layer fibers of a biaxial structure, the droplet can "restore" CB state (see Video S3, Supporting Information).To highlight this special phenomenon, we denote this Cassie-Baxter-"restoring" process as CaRe wetting.This CaRe wetting is achieved on the biaxial structure because i) the overall greater energy barrier to W state compared to that for uniaxial structure hinders the droplet from further wetting the lower-layer fibers, and ii) the local minimum E iiʹ for the metastable State(iiʹ) is >E CB (∆E iiʹ−CB = 9.7), which allows the droplet to restore CB state for being thermodynamically more stable.It should be noted that, in addition to the biaxial structure, the porous surface of fibers also contributes to realizing the CaRe wetting because the pores in fibers ensure that the droplet still contacts vapor phase when the droplet touches the lower layer fibers, which elevates E iiʹ and makes State(iiʹ) less stable than the CB state.
In addition to uncovering the mechanism behind the observed CB state on the biaxial structure (see Figure 2a) through the CaRe wetting, we further investigated the impact of the structural parameters on the CaRe wetting, aiming at providing a guidance for structure optimization.The impact was evaluated by the energy difference between the CB state and State(iiʹ), i.e., ∆E iiʹ−CB ; a higher ∆E iiʹ−CB facilitated the CaRe wetting and also resulted in a more stable CB state.Figure 3c shows a positive correlation between the increase in fiber diameter (d) and the decrease in inter-fiber distance (l) of the upper-layer with ∆E iiʹ−CB .In detail, the normalized ∆E iiʹ−CB increases 19.3 times when increasing the normalized upper-layer d from 0.8 to 2 while l = 7; further decreasing l from 7 to 5 results in an additional 62.3% increase in the normalized ∆E iiʹ−CB .This is because increasing d or decreasing l of the upper-layer fibers requires greater deformation and thereby more energy for the droplet to wet the lower-layer fibers.
Figure 3d shows the variations of ∆E iiʹ−CB with changing lowerlayer d and l.The normalized ∆E iiʹ−CB decreases dramatically by 94.3% when increasing the normalized lower-layer d from 0.8 to 3.2 while l = 6.Because a larger lower-layer fiber diameter means a smaller curvature of the liquid-fiber interface, allowing the droplet to wet the lower-layer fibers with smaller deformation; as a result, the E iiʹ is reduced.Moreover, the normalized ∆E iiʹ−CB increases by 51.0% when increasing the normalized lower-layer l from 3 to 8 while d = 0.8; but further increasing l from 8 to 14 leads to a small ∆E iiʹ−CB variation of less than 4.0%.The change in ∆E iiʹ−CB with lower-layer l is attributed to the interference of  [4a-c,10,23] The data markers in purple and grey are the contact angle and roll-off angle from earlier works, respectively.The adopted biaxial structure has f between 0.053 and 0.047.liquid-fiber interfaces among adjacent lower-layer fibers caused by surface tension of liquid.For the lower-layer l in the range of 3-8, the interference among adjacent liquid-fiber interfaces is reduced with increasing l, resulting in a more stable CB state that is reflected by the increase in ∆E iiʹ−CB .However, for lower-layer l >8, the interference is further reduced, leading to an insignificant effect of l on ∆E iiʹ−CB .
In summary, the energy profile in Figure 3b unveiled the CaRe wetting which explains the observed CB state for the biaxial structure in experiment whereas the uniaxial structure reaches W state (see Figure 2a).The occurrence of CaRe wetting is related to the structural parameters of the biaxially aligned porous fibers: increasing the upper-layer d (from 0.8 to 2), lower-layer l (from 3 to 8), and decreasing the upper-layer l (from 7 to 5), lower-layer d (from 3.2 to 0.8) facilitates the CaRe wetting and results in a stable CB state.In the current study, the CaRe wetting is considered the key advantage of the biaxial structure which enables us to further adjust the structural parameters to strategically approach superhydrophobicity.

Structure toward Superhydrophobicity
This section follows the guidance from Figure 3c,d to construct a biaxial structure that facilitates CaRe wetting and adjusts the structure parameters through experiments to achieve superhydrophobicity.Considering the practical fabricating range of electrospinning and the preceding guidance, we constructed the upper-layer fibers at 8.1 ± 0.62 μm in diameter, while the lowerlayer fibers were smaller in diameter at 3.95 ± 0.25 μm.This combination of fiber diameters can promote the CaRe wetting, enabling us to strategically approach superhydrophobicity by reducing structure solidity.The CaRe wetting is crucial for attaining superhydrophobicity because the CaRe wetting ensures droplet only contacting upper-layer fibers.Thus reducing the structure solidity (by increasing pore size and average inter-fiber distance of the upper-layer fibers) also decreases the liquid-solid contact fraction f, which is the ratio of the liquid-solid interface area to the plane geometrical area parallel to the substrate, [16] see Equation S6.2 (Supporting Information).The hydrophobicity is then improved with decreasing f as preceding explained.
Figure 4a shows the measured  c and  r on the biaxial structure with decreasing f.The decrease of f from 0.187 to 0.101 in the yellow shaded area was attributed to the increase of average pore sizes from 77 ± 16 to 180 ± 49 nm on fiber-surface (while l was kept at 32.0 ± 3.5 μm).The hydrophobicity of the biaxial structure was improved with increasing pore size as evidenced by the increase of  c from 141.0 ± 0.4°to 157.3 ± 0.4°and the decrease of  r from 48.0 ± 0.4°to 28.4 ± 0.8°.In addition, the upper-layer l was further adjusted with the largest pore size of 180 ± 49 nm in fibers.As shown in the blue shaded area of Figure 4a, reducing f from 0.101 to 0.047 by increasing upper-layer l further improved hydrophobicity.The fiber-dressed surface became superhydrophobic when f was below 0.053, which corresponded to a upper-layer l of 81.5 μm.Therefore, the adopted biaxial structure for further studies had f values between 0.053 and 0.047.Further decreasing f below 0.047 by increasing the upper-layer l was not adopted because keeping the upper-layer fibers in a closer distance facilitated the CaRe wetting as shown in Figure 3c.
In contrast to previous studies of developing electrospun fiberbased superhydrophobic structures that used a single droplet volume to measure  c and  r (or a larger droplet volume to measure  r ), the current study evaluated the hydrophobicity for a range of droplets between 3 and 10 μL without any bias (see Figure 4b).This approach provides an objective evaluation of the superhydrophobicity of the biaxial structure and improves the reliability of the results.10a] The results in Figure 4b show that the effect of droplet volume on  c was negligible because  c maintained relatively stable between 159.3 ± 0.9°and 161.5 ± 0.7°with varying droplet volumes.The variation of 2.2°is close to the inherent error of sessile drop method for unstructured surfaces (i.e., ≈±2°), [24] additional minor error may result from the surface heterogeneity of the fibers used in the experiment.In contrast,  r decreased from 10 ± 1°t o 3 ± 0.6°with increasing droplet volume from 3 to 10 μL.This observed trend agrees with the theoretical relationship between droplet volume (V) and  r , which suggests that sin  r is proportional to V −2/3 . [25]In addition, the contact angle and roll-off angle along the direction perpendicular to the upper-layer fibers were 153.6°± 0.7°and 9°± 0.6°, respectively, so the resulting structure achieved superhydrophobicity in both directions.It is worth emphasizing that the method of electrospinning often fails in developing superhydrophobic surfaces without extra fiber surface modification.Previous studies typically involve complex fabrication processes and use multiple raw materials.For example, preparing fluorinated polymers, blending additives into electrospinning solutions, and post-treatments of the electrospun fibers are often required (see Table S3, Supporting Information) by others.However, we produce a superhydrophobic fibrous structure with a simple electrospinning using only one commercial polymer.More importantly, the developed structure possesses a low solidity owing to its sparse fiber arrangement inspired by the ancient Chinese Plain Gauze Gown, which further enables a wide range of applications as follows.

Waterproof Breathable and Superhydrophobic Membrane
In the current study, our biaxial structure of aligned porous fibers provides an opportunity to address the challenge of developing a simple approach for fabricating waterproof, breathable, and superhydrophobic (WBS) membranes.1b] Superhydrophobicity is an additional desired feature for waterproof and breathable membranes for self-cleaning effect. [26]Regarding fabrication, electrospinning is widely adopted for developing waterproof and breathable membranes because the electrospun fibrous structure provides a good basis for breathability. [27]However, despite the substantial progress made in this field, existing methods only achieve two or three out of four features of waterproofing, breathability, superhydrophobicity, and simplicity (see Table S3, Supporting Information).The demonstration of our approach that achieves all four features is shown as follows.
Figure 5a-c and Video S4 (Supporting Information) demonstrate the application of the biaxially aligned porous fibers for developing WBS membrane.The fibers were dressed on a commercial nylon mesh that is originally water permeable.Remarkably, despite the developed biaxially aligned porous fibers being sparsely arranged with only two layers of fibers, they transformed the mesh into a waterproof membrane.As shown in Figure 5a, water passed the pristine mesh and entered the bottom vessel, whereas the fiber-dressed mesh can hold the water in the top vessel.
In addition, the breathability of the fiber-dressed mesh was demonstrated in Figure 5b; the water vapor generated by heating passed the fiber-dressed mesh and condensed in the upper vessel.The breathability was then quantified by a water vapor transmission rate (WVTR) of 20.8 ± 0.1 kg m −2 d −1 .To the best of our knowledge, the WVTR herein is the highest value reported so far for electrospun WBS membranes (see Table S3, Supporting Information) due to its low solidity.Moreover, the biaxial structure also imparted superhydrophobicity to the mesh, as demonstrated by the rolling of a droplet from its surface (see Figure 5c and Video S4, Supporting Information).
It should be noted that the nylon mesh substrate used herein is only for demonstration.The developed fibrous biaxial structure is essentially a versatile dressing layer that can be applied on various substrates, such as conventional electrospun membrane, microporous media, and metal mesh.The choice of substrate depends on mechanical strength, hydrostatic pressure, breathability, and so on, in different applications.For example, Figure 5d and Video S5 (Supporting Information) demonstrate the modification of an electrospun poly(vinylidene fluoride-cohexafluoropropylene) (PVDF-HFP) membrane.Prior to modification, droplets were pinned on the surface of the pristine PVDF-HFP membrane, even by placing the membrane vertically.In contrast, after dressing the membrane with the biaxially aligned porous fibers, droplet rapidly rolled off the surface.

Surface Layer for Wearable Electronics
The emerging wearable electronics have advanced considerably for healthcare, energy harvesting, human-machine interface. [28]o improve user comfort and device longevity, it is important to incorporate a breathable and superhydrophobic surface layer into wearable electronics.The surface layer enables sweat evaporation, preventing skin irritation and inflammation from prolonged use, and also protecting the electronics from wetting damage.This section demonstrated such an application, where a flexible nylon web, representing an electronic layer, was patched onto a human forearm with a surface layer of biaxially aligned porous fibers.The temperature change of the skin area after being covered for 60 min reflects the impact of the covering material on sweat evaporation because a material with inadequate breathability would cause a temperature drop due to the evaporation of the accumulated sweat (see Figure 6a).As shown in Figure 6b, the skin area showed unnoticeable change in temperature after being covered by the biaxially aligned porous fibers, indicating sufficient breathability for wearable uses.In addition, rapid rolling of a water droplet on the same biaxial structure indicated its superhydrophobicity (see Figure 6c and Video S6, Supporting Information), if used as a surface layer for wearable electronics on human skin.This feature can be particularly useful in protecting devices that are sensitive to water intrusion, such as the wearable nano-generator.In addition, the developed surface layer was semi-transparent with a light transmittance of 83.6 ± 3.1%.The transparency was demonstrated in Figure 6d, where the pattern under the fiber-dressed glass is visible.The transparency of the developed surface layer is useful for allowing direct observation and close monitoring of wearable electronics.

Conclusion
Inspired by a Chinese cultural relic known as Plain Gauze Gown, we demonstrated the development of a superhydrophobic and highly breathable structure using biaxially aligned porous fibers with sparse average inter-fiber distance.The developed structure exhibited an unexpectedly stable Cassie-Baxter (CB) wetting state that supported droplets staying on its upper-layer fibers.Further investigations on the free energy profiles of droplet wetting fibrous structures unveiled a unique CaRe wetting that explains the unexpected wetting behavior.The results indicate that the biaxial structure has a relatively greater energy barrier that hinders wetting transition to Wenzel state; additionally, the CB state is thermodynamically more stable compared to the wetting state of droplet contacting lower-layer fibers.Then, this CaRe wetting enabled us to achieve superhydrophobicity by rationally reducing the structure solidity.
Remarkably, the developed superhydrophobic structure is not only breathable, but also waterproof and semi-transparent.These characteristics make the developed structure suitable for a wide range of potential applications, including developing WBS membranes with simple preparation and low cost, and as a surface layer for wearable electronics that allows quick evaporation of sweat and prevents water intrusion.

Experimental Section
Fabrication of Aligned Porous Fibers: The aligned porous were fabricated by electrospinning as follows.First, a polymer solution was prepared by dissolving 1.9 g of polystyrene (PS, obtained from Sigma-Aldrich, US, with a molecular weight of 280 kDa and an intrinsic water contact angle of 95°) into 10 mL of tetrahydrofuran (THF, obtained from Sigma-Aldrich, US, ACS reagent grade ≥ 99%).Then, the solution was loaded into a glass syringe with a 22-gauge metallic needle for electrospinning with solution feeding rate of 4.4 ml h −1 , applied voltage of 6 kV, and needle tip to collector distance of 10 cm.A rotating fin collector covered by transparent tapes was used to collect the uniaxially aligned fibers.The collector rotated at a speed of 450-1,000 r min −1 .The electrospinning duration varied from 30-270 s to adjust the average inter-fiber distance, see Section S5 (Supporting Information) for the relationship between the electrospinning duration and average inter-fiber distance.The transparent tape has an intrinsic water contact angle of 95°.To create pores in the fibers, the relative humidity (RH) inside the electrospinning chamber was controlled between 40% and 80% using a portable humidifier (S09217, Sparoom, US).Finally, the biaxially aligned porous fibers were produced by rotating the tape 90°f or depositing upper-layer fibers.
Fiber Morphology Characterization: The morphology of the prepared electrospun fibers was characterized using a field-emission scanning electron microscope (SEM, Zeiss Leo 1530, Germany) at 5.0 kV.SEM images at high magnifications of 10-20kX were analyzed using ImageJ software to measure the average fiber diameter, average pore size, and the pore area fraction on the produced fibers.In addition, an optical microscope with a digital in-lens camera (Swift SW380B, USA) was used to capture images of the produced fibers at low magnifications of 40-1,000X.These images were then analyzed using Swift Imaging 3.0 software to determine the average center-to-center inter-fiber distance.
Measurements of Hydrophobicity, Breathability, and Transparency: The contact angle ( c ) and roll-off angle ( r ) of the fiber-dressed surfaces were measured using the sessile drop method with a goniometer (Model 250, ramé-hart instrument, USA).Water droplets with a volume of 3 μL, unless otherwise specified, were generated by an automated dispensing system of the goniometer and deposited onto the testing surfaces.The contact angles along the fiber direction were measured from the images captured by the ramé-hart DROPimage Advanced software.After that, the roll-off angle was determined by tilting the sample platform of the goniometer in the direction that droplets roll along the fiber direction.
The breathability of samples was measured by the water vapor transmission rate (WVTR) following the ASTM E96 standard.In a typical test, CaCl 2 desiccant was sealed by the sample in a testing cup (VF2201, TQC Sheen, USA) with 25 cm 2 opening, which was placed in an environmental chamber (SH-222, Espec, Japan) controlled at 38 °C and 90% RH.The weight of the testing cup was measured every 10 min to calculate the WVTR Using Equation (2).
where m f and m i are the final and initial weights of the testing cup, respectively.Δt is the test period of time, and A is the cup opening area.The transparency of the sample was quantified by the light transmittance at visible light wavelength of 380-760 nm.The measurement was carried out using a UV-vis spectrometer (Lambda 1050, PerkinElmer, USA).All measurements of hydrophobicity, breathability, and transparency were repeated 3-5 times, and average values were presented herein with standard deviation.
Model Development: As shown in Equation (1), the free energy (E) of the water droplet was essentially related to the contacting areas between liquid and vapor (A LV ), liquid and fiber (A LF ), and liquid and substrate (A LS ).The latter three were variables that change with the droplet penetrating fiber layer and spreading on the fibrous surface.To account for the deformation of contacting interfaces, especially in cases where the droplet was in contact with biaxially aligned fibers, the analytical solution of the Equation (1) became challenging because of the complex geometry.Therefore, this study employed the Surface Evolver (SE, version 2.70) finite element code to numerically solve this equation and simulate the wetting state by minimizing the total interfacial energy.The initial configuration of a water droplet on a fiber-dressed surface was shown in Section S6 (Supporting Information).The effect of gravity on droplet deformation was negligible because the radius of a 3 μL droplet (0.89 mm) was less than the capillary length of 2 mm. [29]The effect of porous fibers on wetting was approximately considered equivalent to that of using smooth fibers with elevated  Y,F in the modeling because the SE was incapable of adding porous textures to fibers.In detail, the  Y,F was amended to 123°for fiber-surface with pores produced at 60% RH instead of using the original intrinsic contact angle (i.e., 95°) of PS material in the SE modeling (see Section S7, Supporting Information).In addition, the length and energy related parameters in the model were normalized to their dimensionless forms with respect to d c , and  LV d c 2 , [30] respectively.d c was set as 5 μm to represent the characteristic diameter of the produced fibers.
Ethical Statment: Informed written consent was obtained from the participant prior to participation in the research.

Figure 1 .
Figure 1.Fabrication of sparsely aligned porous fibers by electrospinning for investigating wetting behavior: a) Plain Gauze Gown with optical microscope image of the fibrous structure demonstrates the inspiration for fabricating electrospun fibers with an average center-to-center inter-fiber distance of l and fiber diameter of d.The images of the Plain Gauze Gown are obtained from Hunan Museum, China.b) Scanning electron microscope (SEM) and optical microscope images of electrospun fibers produced at 40%-80% RH, showing the porous fiber-surfaces with different pore sizes (o) and two alignment patterns.c) Illustration of the water contact angle ( c ) and roll-off angle ( r ) measurements on a fiber-dressed surface.

Figure 2 .
Figure 2. Water droplets with 3 μL in volume wetting surfaces dressed by sparsely aligned porous fibers.a) Contact angle and roll-off angle on surfaces dressed by uniaxially or biaxially aligned porous fibers with an average fiber diameter of 3.95±0.25μm, an average pore size of 160 ± 31 nm, and pore area fraction of 0.5 in fiber-surface.Most error bars are smaller than the size of the data markers.b) Images of the final seconds of droplet free evaporation on the biaxially aligned porous fibers with droplet diameter decreased to <300 μm.The upper-layer average inter-fiber distance of the droplet covered area is 21.6 μm.The evaporation is under room temperature of 22 °C and RH of 24%.The yellow dot-lines highlight liquid edges in interact with the upper-layer fibers, and the red dot-lines highlight the edges in interact with the lower-layer fibers.c) Experimental data points mapped in a wetting state chart obtained from modeling the uniaxially aligned fibrous structure with a fiber diameter of 3.95 μm.The droplet width is measured from the direction along to fibers.

Figure 3 .
Figure 3. Free energy of water droplet wetting surfaces dressed by sparsely aligned porous fibers.The droplet volume, inter-fiber distance, fiber diameter, and pore area fraction were 3 μL, 30 μm, 4 μm, and 0.5, respectively.a) Illustration of four different wetting states.b) Energy profile for a water droplet wetting uniaxial structure and biaxial structure, the insets are the wetting states with a local minimum free energy obtained from modeling.c) Normalized ∆E iiʹ−CB versus upper-layer fiber diameter and inter-fiber distance of the biaxial structure.Low layer normalized fiber diameter and inter-fiber distance were set as 0.8 and 6, respectively.d) Normalized ∆E iiʹ−CB versus lower-layer fiber diameter and inter-fiber distance of the biaxial structure.Upper layer normalized fiber diameter and inter-fiber distance were set as 0.8 and 6, respectively.The normalization of energy and all length related parameters were with respect to  LV d c 2 , and d c , respectively, where d c was 5 μm to represent the characteristic diameter of the produced fibers.

Figure 4 .
Figure 4. Adjusting the structure of biaxially aligned porous fibers to strategically approach superhydrophobicity. a) Effect of upper-layer pore size and average inter-fiber distance on the measured water contact and roll-off angles.The fiber diameters for the upper-and lower-layer are 8.1 ± 0.62 and 3.95 ± 0.25 μm, respectively.The average inter-fiber distance for both layers in the yellow shaded area and the lower-layer in the blue shaded area are 32.0 ± 3.5 μm.Most error bars are smaller than the size of data markers.b) Variation of surface hydrophobicity for different droplet volumes on the biaxial structure, and the comparison to other electrospun fiber-based superhydrophobic surfaces reported in literature.[4a-c,10,23]The data markers in purple and grey are the contact angle and roll-off angle from earlier works, respectively.The adopted biaxial structure has f between 0.053 and 0.047.

Figure 5 .
Figure 5. Demonstration of potential application for developing waterproof breathable and superhydrophobic membrane.a) Applying biaxially aligned porous fibers on a mesh for waterproofing.Insets are the optical microscopic images of the mesh before and after dressing by the fibers.b) Breathability of the fiber dressed mesh.c) Superhydrophobicity of the fiber dressed mesh.The water droplets are dyed in orange.d) Turning an electrospun PVDF-HFP membrane from droplet-pinning to superhydrophobic by dressing the fibers.The water droplets are dyed in orange.The white arrow is the direction of gravity.

Figure 6 .
Figure 6.Demonstration of potential application as the surface layer for protecting wearable electronics.Thermal images of patching a) a non-breathable adhesive tape and b) the developed biaxially aligned porous fibers on human forearm skin for 60 min and removal afterward.c) Superhydrophobicity of forearm skin area covered by the biaxially aligned porous fibers.Water droplet is in orange.d) Transparency of the developed biaxially aligned porous fibers.