Janus membrane with novel directional water transport capacity for efficient atmospheric water capture

Baona Ren , Haohong Pi , Xin Zhao , Miaomiao Hu , Xiuqin Zhang , Rui Wang and Jing Wu *
Beijing Key Laboratory of Clothing Materials R & D and Assessment, Beijing Engineering Research Center of Textile Nanofiber, School of Materials Design & Engineering, Beijing Institute of Fashion Technology, Beijing 100029, China. E-mail: a.wujing@163.com

Received 19th February 2021 , Accepted 29th April 2021

First published on 30th April 2021


Abstract

Fresh water scarcity has become a crisis affecting human survival and development. Atmospheric water capture with remarkable advantages such as energy-independence and low-cost is supposed to be a promising way to address the problem. Herein, a facile strategy is presented to design a membrane material with efficient atmospheric water capture capacity and high practical significancy. A hybrid Janus membrane with anisotropic wettability and morphology is fabricated by integrating electrospinning and in situ surface oxidation methods. Taking advantage of the anisotropic wettability and strong force provided by directional wicking to draw water drops from a hydrophobic to a hydrophilic layer, the Janus membrane exhibits novel directional water droplet transport and possesses efficient and excellent atmospheric water capture capacity. Janus membrane with larger pores in the hydrophobic layer shows higher atmospheric water capture capacity than that with smaller pores. Furthermore, the hybrid Janus membrane is successfully implemented in soil water retention in the plant cultivation process. This work provides an insight into the facile design of the Janus membrane for fresh water capture, which is important to extend its practical applications.


Introduction

In some parts of the world, water resources, especially fresh water, is severely scarce due to population boom, overexploitation and the neglect of protection in the process of rapid development in industry and society. Such water crisis affects the natural ecological security and even jeopardizes the survival of people all over the world.1–4 Until now, a series of methods and technologies including wastewater secondary processing5 and seawater desalination6–8 have been applied to alleviate and solve the problem. Some of them are indeed effective, but they also show disadvantages, such as energy-dependence, high-cost, and other non-renewable resource waste. As we know, atmospheric water, which constitutes approximately 10% of all fresh water on earth, is a significant but largely untapped fresh water source.9,10 If atmospheric water can be captured efficiently, it may provide a promising and efficient way to get fresh water due to its high content and widespread existence in nature.

Plants and animals, after long-time evaluation, have gained unique water capture capabilities for their survival, especially in regions devoid of water. For example, the Namib Desert beetles collect dew using bumpy alternating hydrophilic and hydrophobic patterns on their body surface. Tiny droplets are captured on the hydrophilic regions and transported to the hydrophobic areas.11–18 Spider silk and drought-tolerant cactus spines taking advantage of the synergetic effect of both the shape and wettability gradient along a periodic spindle-knot structure and conical spines, respectively, capture water droplets and transport them directionally.19,20 Taking inspiration from these excellent “water captors” in nature, which elucidate the critical role of wettability irregularity and hierarchical structure gradient in water transport and capture, biomimetic 1D filaments and 2D surfaces have been so far designed and made for dew/atmospheric water capture.21–28 Further in-depth studies will be focused on facing the challenge of designing materials that can improve the water capture efficiency significantly. In recent years, the Janus membrane with the typical feature of a hydrophobic (superhydrophobic) layer on one side and a hydrophilic (superhydrophilic) layer on the other side has been used as a “water diode”, viz, water can transport from the hydrophobic (superhydrophobic) layer to the hydrophilic (superhydrophilic) layer but flow in the inverse direction is blocked, and has attracted much attention and become a research hotspot in many realms.29–33 A wide variety of strategies have been proposed to fabricate the Janus membrane including surface grafting,34 self-assembly,35 atomic layer deposition,36 and femtosecond laser technology.37 However, understanding on how to obtain the Janus membrane conveniently and apply it to improve the water capture efficiency, bestowing it with high practical significance, is still immature thus far.

Herein, we demonstrate a facile strategy to develop a hydrophobic/hydrophilic hybrid Janus membrane with anisotropic wettability and morphology by integrating flexible electrospinning and in situ surface oxidation methods. The hydrophilic hierarchical copper hydroxide (Cu(OH)2) layer was obtained through in situ surface oxidation, and selected as the substrate. The hydrophobic nanofibrous layer was constructed on it via electrospinning poly(vinylidene fluoride-co-hexafluoropropylene) mixed with 1H,1H,2H,2H-perfluor-odecyltrimethoxysilane (PVDF-HFP/F). By virtue of the synergistic effect of the anisotropic wettability and porous structure, the Janus membrane exhibited novel directional water transport and excellent atmospheric water capture capacities. Such a Janus membrane opens a new pathway for preserving soil moisture, promoting plant growth and relieving agricultural and fresh water scarcity. Meanwhile, the facile strategy offers advantages such as mild reaction conditions, convenient operation, and the possibility of large-scale production in the future.

Experimental section

Materials

Poly(vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP, Mw = 455[thin space (1/6-em)]000) was bought from Sigma-Aldrich (St Louis, MO, USA). N,N-Dimethylformamide (DMF), ethanol and acetone were purchased from Beijing Yili Fine Chemical Co. (Beijing, China). 1H,1H,2H,2H-Perfluorodecyltrimethoxysilane was obtained from J&K Scientific Ltd. Copper mesh wire (150 mesh), sodium hydroxide (NaOH) and potassium peroxydisulfate (K2S2O8) were provided by Beijing Chemical Works (Beijing China). Deionized water was self-made in the laboratory. All the chemicals were of analytical grade and used as received.

Preparation of the Janus membranes

Fabrication of Cu(OH)2 mesh wire with hierarchical nanosheet structure. A copper (Cu) mesh wire (length × width = 4 cm × 3 cm) was used. To remove the possible surface contaminants, the Cu mesh wire was first cleaned in acetone ultrasonically, then successively rinsed in ethanol and deionized water with ultrasonication for a certain time. Finally, it was dried at 60 °C in a vacuum oven for further use. Such a pre-cleaned Cu mesh wire was immersed into a solution containing 2.5 M NaOH and 0.1 M K2S2O8 for about 20 min at room temperature. Subsequently, the Cu mesh wire was taken out and washed with deionized water. A copper hydroxide (Cu(OH)2) mesh wire with hierarchical nanosheet structure was, thus, successfully obtained. The chemical reaction equation can be expressed as follows:
Cu + 2NaOH + K2S2O8 = Cu(OH)2 + Na2SO4 + K2SO4
Preparation of PVDF-HFP/F electrospun fibrous membrane and PVDF-HFP/F/Cu(OH)2 Janus membrane. The hydrophobic PVDF-HFP/F nanofibers were prepared via electrospinning. Dried PVDF-HFP powder was dissolved in DMF and acetone solvent (weight ratio = 7[thin space (1/6-em)]:[thin space (1/6-em)]3) under magnetic stirring at 40 °C for 8 h to form 15 wt%, 17.5 wt%, 20 wt%, 22.5 wt% and 25 wt% electrospun precursor solutions. Furthermore, 1H,1H,2H,2H-perfluorodecyltrimethoxysilane (F) (PVDF-HFP[thin space (1/6-em)]:[thin space (1/6-em)]F = 10[thin space (1/6-em)]:[thin space (1/6-em)]1 w/w) was added to the PVDF-HFP electrospun solution in order to adjust the wettability of the fibrous membranes (the fibrous membrane was named as PVDF-HFP/F). The distance and applied voltage were set as 15 cm and 10–15 kV, respectively. The structure of the fibrous membrane could be controlled by changing the electrospun solution concentration. During the electrospinning process, the Cu(OH)2 mesh wire was fixed to the drum (rotation speed = 200 rpm) as the collector, and acted as the hydrophilic layer of the anisotropic wettability Janus membrane. After electrospinning, the PVDF-HFP/F/Cu(OH)2 Janus membranes were successfully obtained.

Instruments and characterization

The surface morphologies of the pristine Cu mesh wire, Cu(OH)2 mesh wire, electrospun PVDF-HFP/F fibrous membrane and Janus membrane were observed via a scanning electron microscope (SEM, JSM-6700, Tokyo, Japan). The fiber diameters and pore size distributions of the above-mentioned samples were analysed by using a professional image analysis software (provided by the SEM system, Phenom ProX). Water contact angles were tested by using a water contact angle measurement system (OCA 20, Dataphysics, Germany). Deionized water droplets (5 μL) were dripped onto the membrane surface. The average water contact angle values could be obtained by measuring at six different positions of the same sample. The interaction between water and the membrane was tested by a high-sensitivity micro-electro-mechanical balance system (Jingong, Shanghai, China). Two testing types were employed. For Type 1: a water droplet was suspended and fixed in a metal loop. The membrane moved upward with a constant speed of 0.005 mm s−1 to contact the water droplet. Once the membrane surface was in contact with the water droplet, it began to move downward immediately. At this time, the water droplet was dragged by the force provided by the surface. Once the water droplet was released or separated from the surface during the downward moving process, the surface stopped moving. The moving distance and the force change during the process were recorded. Thus, the distance–adhesive force curve was obtained. For Type 2, the water droplet contacted the surface, and remained in contact with the surface rather than the surface or water droplet moving. Once the water droplet penetrated through the membrane, the adhesive force changed. When the water droplet was fully penetrated, the force finally dropped to zero. The time and the force change were recorded, according to which the time–adhesive force curve was obtained. The crystal structures of the pristine Cu mesh wire and Cu(OH)2 mesh wire were examined by a Rigaku-D/max 4000 V X-ray diffractometer equipped with Cu Kα radiation (λ = 0.15418 nm) at a step width of 5° min−1 (XRD, Bruker, Karlsruhe, Germany). The chemical compositions and elemental distribution maps were tested using energy-dispersive X-ray spectroscopy (EDS, Quanta 650FEG, FEI). Fourier transform infrared (FTIR) spectra of the as-prepared samples were obtained by a Nicolet 8700 FT-IR spectrometer (Madison, WI, USA) in the wave range from 400 cm−1 to 4000 cm−1. The thickness gauge (CHY-U, Jinan, China) was used to measure the thicknesses of the hybrid Janus membranes.

Atmospheric water capture

The atmospheric water capture capacities of the as-prepared samples, including the pristine Cu mesh wire, Cu(OH)2 mesh wire, PVDF-HFP/F-290, PVDF-HFP/F-810, PVDF-HFP/F-810/Cu(OH)2 and PVDF-HFP/F-290/Cu(OH)2 membranes were evaluated by using a water trapping system. A humidifier (Air Humidifier Ultrasonic Cool Mist Steam, China) was used to generate the water droplets at a velocity of 70 cm s−1. All the samples were cut into pieces of 4 cm × 3 cm (length × width), and were fixed on a holder. Meanwhile, the distance between the atmospheric water and the membrane surface was fixed and the angle was adjustable. After being exposed to tiny water droplet flow for a certain time, the samples were weighed. Finally, the collected water was utilized to evaluate the atmospheric water capture capacity of the membranes.

Plant cultivation in lab

Plant cultivation was performed in two groups in the lab to explore the influence of the directional water transport behaviour of the PVDF-HFP/F/Cu(OH)2 Janus membrane on plant growth qualitatively. Mint was selected as the model plant. For group I, a PVDF-HFP/F-810/Cu(OH)2 (length × width = 4 cm × 3 cm) Janus membrane was placed on the soil (hydrophobic layer upward) where mint grew. For group II, as a control group, there was no Janus membrane placed on the soil. In the experimental process, for group I and II, we regularly supplied moisture through a humidifier at constant temperature (25 °C) and humidity (40%) for two hours every day to simulate atmospheric water environment. The growth of mint was observed and recorded by taking photos every day at the same time.

Results and discussion

The fabrication process of the hybrid Janus membrane with anisotropic wettability and morphology is schematically illustrated in Fig. 1a. A copper (Cu) mesh wire was used. After in situ surface oxidation reaction, copper hydroxide (Cu(OH)2) nanosheets grew from the surface of the Cu mesh wire and fully covered it, forming a hierarchical Cu(OH)2 mesh wire.38 Subsequently, the as-prepared Cu(OH)2 mesh wire was selected as the substrate, and the poly(vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP) mixed with 1H,1H,2H,2H-perfluorodecyltrimethoxysilane (F) (PVDF-HFP[thin space (1/6-em)]:[thin space (1/6-em)]F weight ratio = 10[thin space (1/6-em)]:[thin space (1/6-em)]1) fibrous membrane (PVDF-HFP/F) was electrospun on it. The scanning electron microscopy (SEM) images of both the pristine Cu and Cu(OH)2 mesh wires are shown in Fig. 1b and c, respectively. According to the high-magnification SEM images, it is obvious that the surface of pristine Cu was smooth (insert, Fig. 1b), while, after the in situ surface oxidation reaction, a layer of nanosheets was covered on the Cu mesh wire (insert, Fig. 1c). X-ray diffraction (XRD) and Fourier transform infrared (FTIR) spectra were analysed to deeply verify the composition of the as-treated Cu mesh wire (Fig. S1). The result confirmed that the nanosheets were orthorhombic-phase Cu(OH)2 crystals, which was in accordance with the values in the standard card (JCPDS Card No. 13-420).39 After the PVDF-HFP/F nanofibers were electrospun on the Cu(OH)2 mesh wire (Fig. 1d), a free-standing Janus membrane (PVDF-HFP/F/Cu(OH)2) consisting of two seamless layers was successfully fabricated (Fig. 1e).
image file: d1nr01120k-f1.tif
Fig. 1 (a) Schematic of the hybrid Janus membrane fabrication process. SEM images of (b) the pristine Cu mesh wire (insert: partial magnified image); (c) Cu(OH)2 mesh wire with hierarchical nanosheet structure (insert: partial magnified image); (d) the Janus membrane viewed from the top of the PVDF-HFP/F fibrous membrane side. (e) SEM image of the cross-section of the PVDF-HFP/F-810/Cu(OH)2 Janus membrane. (f and g) SEM images, pore size distributions and statistic average fiber diameters (insert) of the electrospun PVDF-HFP/F-290 and PVDF-HFP/F-810 fibrous membranes. EDS element distribution maps of the partial enlarged Janus membrane from the PVDF-HFP/F fibrous side (h) and the Cu(OH)2 mesh wire side (i).

To explore the influence of membrane morphology on atmospheric water capture, the pore size of the electrospun PVDF-HFP/F fibrous membrane was regulated by changing the electrospun solution concentration, electrospun injection velocity as well as spinneret inner diameter, which have been proved to have important effects on fiber morphology in our previous works.30,40–43 In this work, we found that the electrospun solution concentration had a major influence on the fiber morphology. Smooth PVDF-HFP/F fibers with the average diameter from about 130 nm to 810 nm could be obtained when the electrospun solution concentration was in the range from 15 wt% to 25 wt% (more details can be seen in ESI Fig. S2). Fig. 1f and g are the SEM images of the PVDF-HFP/F fibrous membranes with great difference in average fiber diameter, i.e., when the PVDF-HFP/F electrospun concentration was 17.5 wt% and the spinneret inner diameter was 1.2 mm, the average fiber diameter was 293 ± 13 nm (insert, Fig. 1f, named as “PVDF-HFP/F-290”), and the PVDF-HFP/F with the average fiber diameter of about 806 ± 29 nm (insert, Fig. 1g, named as “PVDF-HFP/F-810”) was prepared when the electrospun solution concentration was 25 wt% and the spinneret inner diameter was 1.2 mm. Notably, the pore size of the PVDF-HFP/F fibrous membrane changed as the fiber diameter changed. That is, the pore size was increased with the increase in the average fiber diameter. For PVDF-HFP/F-290, the average pore size was 633 ± 19 nm (right image, Fig. 1f), whereas the pore size for PVDF-HFP/F-810 was up to 1.78 ± 0.69 μm (right image, Fig. 1g).

Energy-dispersive X-ray spectroscopy (EDS) was further used to analyse the element distributions of the PVDF-HFP/F/Cu(OH)2 Janus membrane from both its sides. Fig. 1h and i show the EDS elemental maps of C, O, F and Cu from the PVDF-HFP/F fibrous side and the Cu(OH)2 mesh wire side, respectively. The contents of C, O, F and Cu of the PVDF-HFP/F fibrous side were 31.60 wt%, 7.29 wt%, 41.8 wt% and 19.31 wt%, while, they were 0 wt%, 23.85 wt%, 1.95 wt% and 74.20 wt%, from the Cu(OH)2 mesh wire side, respectively, indicating that the PVDF-HFP/F fibers did not pierce into the mesh wire. Meanwhile, Fig. S3 in ESI further confirmed that the hybrid Janus membrane was composited by two layers, one layer was the PVDF-HFP/F fibrous membrane and the other layer with thicker wire was the Cu(OH)2 mesh wire, indicating that the PVDF-HFP/F fibers were stacked on the Cu(OH)2 mesh wire (The details of element contents are listed in Tables S1(a) and S1(b)).

Fig. 2a and b show the wettability of the Janus membrane. After in situ surface oxidation, the water droplet spread on the surface of the Cu(OH)2 mesh wire and the corresponding water contact angle (WCA) was 44.7° (Fig. 2a). Whereas, the water droplet stayed as a sphere and the corresponding WCA was 138.1° when it dripped on the PVDF-HFP/F fibrous layer (Fig. 2b). It indicated the hydrophilicity of the Cu(OH)2 mesh wire layer and the hydrophobicity of the PVDF-HFP/F fibrous membrane layer, giving anisotropic wettability to the Janus membrane successfully. Apart from the wettability, the interaction between water and the membrane, which plays a major role in the water transport capability, was also detected by measuring the drawing force. Here, we provided two ways to do it. Type 1: a water droplet (5 μL) was brought in contact with the membrane and then pulled off, according to which the force between the water droplet and membranes was probed and recorded (schematic on top part of Fig. 2c). Type 2: the water droplet was brought in contact with the membrane without pulling off, and the real-time force change and water penetrating, spreading or adsorbing time were recorded (a brief schematic on top part of Fig. 2d, and detailed schematics in ESI Fig. S4).


image file: d1nr01120k-f2.tif
Fig. 2 (a and b) Photographs of water droplets on the hydrophilic Cu(OH)2 mesh wire side and the hydrophobic PVDF-HFP/F-810 fibrous membrane side, respectively, and the corresponding water contact angles. (c) Illustration of Type 1 and the force–distance curves. Water droplet (5 μL) was brought in contact with the membrane and then pulled off. (d) Illustration of Type 2 and the force–time curves. Water droplet (5 μL) was brought in contact with the membrane without pulling off.

In Fig. 2c, the drawing force (F) increased gradually during the pulling of the water droplet (moving downward distance (D) increased). Once the water droplet separated from the membrane, F decreased abruptly. The F reached the peak value at the time when the water–membrane separated. The F values for the PVDF-HFP/F-810/Cu(OH)2 and PVDF-HFP/F-290/Cu(OH)2 Janus membranes were 226.38 ± 20.21 μN and 194.04 ± 15.83 μN, respectively, which were higher than that of the hydrophobic PVDF-HFP/F-810 (88.20 ± 9.14 μN) and PVDF-HFP/F-290 (106.82 ± 11.53 μN) membranes, and dropped rapidly in a shorter time (insert in Fig. 2c). The F for the hydrophilic Cu(OH)2 mesh wire was 328.30 ± 31.54 μN. In Fig. 2d, for Type 2, there is a great difference between the hydrophobic PVDF-HFP/F membrane and the hydrophilic Cu(OH)2 mesh. For the hydrophobic PVDF-HFP/F-810 and PVDF-HFP/F-290 membranes, on account of nearly no water adsorption over the measuring time, the force–time curves showed no abrupt changes (red and blue dotted straight line in Fig. 2d). The slight tilting down of the curves can be attributed to water evaporation during the measuring process. While, for the hydrophilic Cu(OH)2 mesh wire, the F value reached the peak value in a very short time due to quick water spreading or adsorption. According to the time–force curves of the Janus membrane, the recorded time was 42.83 ± 4.35 s and 48.77 ± 3.57 s for PVDF-HFP/F-810/Cu(OH)2 and PVDF-HFP/F-290/Cu(OH)2 Janus membranes, respectively. Such difference can be attributed to the fact that the larger pores in the hydrophobic layer accelerated water transport, for larger hydrophobic pores provide smaller repellent force.33

Fig. 3 exhibits the water transport performance of Janus membranes. As schematically illustrated in Fig. 3a, water droplets were dripped on the Janus membrane from both its sides. Fig. 3b is the real-time water transport performance of the PVDF-HFP/F-290/Cu(OH)2 (i) and PVDF-HFP/F-810/Cu(OH)2 (ii) Janus membranes. For both PVDF-HFP/F-290/Cu(OH)2 and PVDF-HFP/F-810/Cu(OH)2, when the water droplet dripped on the hydrophobic fibrous side, it could pass through to the hydrophilic Cu(OH)2 side. Nevertheless, when the Janus membrane was overturned, the water droplets spread and blocked on the hydrophilic Cu(OH)2 layer rather than moving spontaneously. That is the directional water transport. Such directional water transport performance was further evaluated by water transport time. For PVDF-HFP/F-810/Cu(OH)2, the time for fully directional water transport was 29 s, which was shorter than that of PVDF-HFP/F-290/Cu(OH)2 (35 s), demonstrating that the directional water transport capacity of PVDF-HFP/F-810/Cu(OH)2 was superior to that of PVDF-HFP/F-290/Cu(OH)2. The Janus membrane that is analogous to a “water-diode”, which has also been elucidated in our previous studies, allows directional water transport, resulting from the anisotropic wettability in the thickness direction of the membrane.30,33 That is, the directional driving force for the Janus membrane was provided by both the repellent force caused by the hydrophobic porous layer and the capillary force provided by the hydrophilic porous layer.


image file: d1nr01120k-f3.tif
Fig. 3 (a) Schematic of the measurement of water transport performance from both sides of the Janus membrane. (b) Snapshots of the water transport phenomenon on (i) the PVDF-HFP/F-290/Cu(OH)2 and (ii) the PVDF-HFP/F-810/Cu(OH)2 Janus membrane. (c) The relationships of electrospinning time, membrane thickness and directional water transport.

When liquid droplets attach to the Janus membrane, the transport modes commonly feature: directional or “one-way” transport, bidirectional transport and bilateral obstruction. It can be interpreted that not all Janus membranes can achieve directional liquid transport, which depends on the thickness distribution of the combined hydrophobic–hydrophilic layer.30,44–46 Herein, the hydrophobic/hydrophilic layouts of the Janus membrane were changed by regulating the thickness of the hydrophobic fibrous layer via the electrospinning time to explore the synergistic effect of directional water transport. As shown in Fig. 3c, the electrospinning time of the PVDF-HFP/F layer ranged from 0 to 180 s. For PVDF-HFP/F-290/Cu(OH)2, when the thickness of the hybrid Janus film was less than 124.65 ± 0.23 μm (the electrospinning time was shorter than 20 s), the Janus membrane showed bidirectional penetration. When the thickness of the hybrid Janus film reached 124.65 ± 0.25 μm (the electrospinning time increased to 30 s), it exhibited directional water transport. Once the thickness of the hybrid Janus film exceeded 125.92 ± 0.17 μm (the electrospinning time was over 120 s), water could not penetrate across both sides of the Janus membrane. Similarly, for PVDF-HFP/F-810/Cu(OH)2, when the thickness of the hybrid Janus film increased from 124.1 to 125.03 ± 0.15 μm (the electrospinning time increased from 0 to 20 s), the Janus membrane exhibited directional water transport. However, if the thickness of the hybrid Janus film exceeded 126.52 ± 0.21 μm (the electrospinning time was over 120 s), the Janus membrane revealed bilateral obstruction behaviour. The detailed relationships of the electrospinning time, membrane thickness and unidirectional water transport capacity are also presented in ESI Table S2. To explain the observed results, it might be considered that (i) if the thickness of the hydrophobic layer is too thin, the resistance to water is relatively low, resulting in water penetration across the Janus membrane from both its sides; (ii) when the thickness of the hydrophobic layer is increased to a moderate range, the Janus membrane owns the directional transport property; (iii) when the thickness of the hydrophobic layer is too thick, its resistance is extremely high, making the Janus membrane an impermeable membrane.46,47

As we know, anisotropic wettability in the thickness direction of the porous membrane is able to generate directional wicking, which is key for directional water transport. Besides water droplet transport, the atmospheric water capture performance was characterized by weighing the moisture being transported and collected on the back side of the membrane. The real-time atmospheric water capture processes of the hydrophobic PVDF-HFP/F-810, hydrophilic (Cu(OH)2) and Janus membrane PVDF-HFP/F-810/Cu(OH)2 are shown in Fig. 4a. When tiny droplets encountered the hydrophobic PVDF-HFP/F-810 fibrous membrane (process (i) in Fig. 4a), they stayed in a hemispherical or spherical shape and pinned tightly on the membrane surface (droplets 1, 2, 3, 4 marked in yellow and red dotted circles in (i), Fig. 4a). On prolonging the capture time, the droplets grew larger, coalesced (droplet 1 + 2, 3 + 4) and kept growing (droplet 1 + 2(↑), 3 + 4(↑)). However, once attached to the hydrophilic Cu(OH)2 mesh wire, these tiny water droplets spread and formed a water film covering on the wire surface (process (ii) in Fig. 4a). For the Janus membrane, when the tiny droplets contacted the hydrophobic fibrous surface, they grew larger (process (iii) in Fig. 4a). With an increase in the number of contacted droplets, they moved to the hydrophilic layer spontaneously, leaving a fresh area to capture more tiny water droplets.


image file: d1nr01120k-f4.tif
Fig. 4 (a) Atmospheric water capture phenomena of (i) the hydrophobic PVDF-HFP/F-810 fibrous membrane; (ii) hydrophilic Cu(OH)2 mesh wire; (iii) Janus PVDF-HFP/F-810/Cu(OH)2 membrane. (b) Moisture trapping rates of the various samples. (c) Moisture trapping rates of the Janus membrane PVDF-HFP/F-290/Cu(OH)2 and (d) the Janus membrane PVDF-HFP/F-810/Cu(OH)2 at the tilt angles of 0°, 45° and 90°. (e) Schematic of the mechanism of capture of tiny water drops.

Then, to characterize the atmospheric water capture quantitatively, different samples including pristine Cu mesh wire, Cu(OH)2 mesh wire, PVDF-HFP/F-290, PVDF-HFP/F-810, PVDF-HFP/F-810/Cu(OH)2 and PVDF-HFP/F-290/Cu(OH)2 Janus membranes were cut into 4 cm × 3 cm (length × width) pieces, and atmospheric water capture capacity was characterized by measuring the moisture trapping rate (schematically illustrated in ESI Fig. S6). According to Fig. 4b, the moisture trapping rate of the pristine copper mesh, Cu(OH)2 mesh wire, PVDF-HFP/F-290 and PVDF-HFP/F-810 were 520.02 ± 30.22, 204.21 ± 66.71, 154.8 ± 53.67 and 317.43 ± 31.23 mg cm−2 h−1, respectively. It can be found that the moisture trapping rate for the hydrophobic membranes (e.g. PVDF-HFP/F-810, Cu mesh wire) was higher than that of the hydrophilic Cu(OH)2 mesh, attributing to the fact that the water movement on the hydrophilic surface is slower than that on the hydrophobic surface due to the larger flow resistance.13,37,48 In comparison, the Janus membranes showed a higher moisture trapping rate on moisture trapping from the hydrophobic side (1162.2 ± 109.06 mg cm−2 h−1 for PVDF-HFP/F-290/Cu(OH)2; 1821.51 ± 56.18 mg cm−2 h−1 for PVDF-HFP/F-810/Cu(OH)2) than that from the hydrophilic side (249.60 ± 62.94 mg cm−2 h−1 for PVDF-HFP/F-290/Cu(OH)2; 676.53 ± 68.75 mg cm−2 h−1 for PVDF-HFP/F-810/Cu(OH)2). Taking these results into consideration, the PVDF-HFP/F-810/Cu(OH)2 Janus membrane with larger pore size in the hydrophobic layer exhibited a higher atmospheric water capture capacity than that of PVDF-HFP/F-290/Cu(OH)2 with smaller hydrophobic pore size, indicating that the larger pore size can accelerate moisture trapping. We also compared the atmospheric water capture capacity of the Janus membrane at different incident angles (θ) between atmospheric water and the membrane surface of 0°, 45° and 90° (Fig. 4c and d). Obviously, for PVDF-HFP/F-290/Cu(OH)2 and PVDF-HFP/F-810/Cu(OH)2, the atmospheric water capture capacity was highest when the tilt angle was 90° (details can be seen in ESI Table S3), which may be owing to the maximum contact area between atmospheric water and the membrane surface when θ equals 90°.49 Furthermore, the water capture capacity of the Janus membrane under different water droplet velocities was tested. It was found that the water capture capacity increased with the increase in water droplet velocity (Fig. S7).

The electrospun fibrous membrane exhibited the classical wicking dynamics analogous to what was described by Washburn considering a porous structure as a series of cylindrical capillaries with an equivalent radius of the capillary structures.50 The mechanism is illustrated in Fig. 4e, tiny water droplets (blue spheres) pinned on the hydrophobic layer and coalesced to form larger droplets (presented as 1 → 1′ and 2′, 3′ → 2′ + 3′). The grey dotted frame shows the zoomed-in view of the force analysis of tiny water droplets. It can be seen that the larger water droplet suffered two opposite forces along the x direction, i.e., in the horizontal direction, the hydro-percussion (HP) and hydrophobic force (HF). HP intends to make water pass through the porous channel, yet HF resists water penetration. Since the PVDF-HFP/F fibrous membrane was hydrophobic with the water contact angle over 130° and the water droplet stayed in the Wenzel–Cassie state, the water drop's volume was maintained and then elevated with the accumulation of tiny water drops, accordingly. Although HF inclines to block the water drop's entrance into the hydrophobic pores, the grown water droplet may stay in the Wenzel state, in which it can be in full contact with the rugged surface. Furthermore, when the penetrative depth reached the junction of the two layers of the Janus membrane, the capillary force (CF) provided by the adjacent hydrophilic Cu(OH)2 mesh wire together with HP facilitated the penetration of the tiny water droplet and made the process continuous, which left active positions on the hydrophobic PVDF-HFP/F layer of the Janus membrane to continue moisture capture.

The Janus membrane with directional atmospheric water capture capacity, by which moisture can be trapped by the hydrophobic layer then transported to the hydrophilic layer, can be an excellent candidate in soil water retention, which could promote plant growth in cultivation. We designed a simulated experiment to investigate the effects of the Janus membrane on plant growth during the plant cultivation process. As shown in Fig. 5, mint was chosen as the model plant. Two groups of experiments (group I and II, Fig. 5a) were carried out at constant temperature (25 °C) and humidity (40%), simultaneously. For group I, the soil was covered by the Janus membrane (length × width = 4 cm × 3 cm) with the hydrophobic PVDF-HFP/F-810 side upward, while, for group II, there was no membrane placed on the soil (Fig. 5a). During the plant cultivation process, no water was provided, except for two hours atmospheric water environment simulation was generated by a humidifier every day. Then, the daily growth in group I and II was recorded with a camera. The results are shown in Fig. 5b and c. It can be found that there was no obvious change in the growth status in group I, which was covered by a piece of Janus membrane on the soil, after 10 days (Fig. 5b). However, for group II, the leaves of mint began to wilt after 6 days and wither after 10 days (Fig. 5c).


image file: d1nr01120k-f5.tif
Fig. 5 Atmospheric water capture system for plant cultivation. (a) Two experimental samples. Mint was chosen as the model plant. For I, a piece of Janus membrane was covered on the soil (hydrophobic layer upward). For II, there was no membrane. (b and c) Images of samples in group I and II, respectively, during 10 days, recorded by a camera at the same time every day.

Conclusions

In summary, we have successfully fabricated a hybrid Janus membrane with anisotropic wettability and morphology through facile electrospinning technology and in situ surface oxidation. On account of the directional water transport capability provided by both the anisotropic wettability and directional wicking to draw water drops from the hydrophobic to the hydrophilic porous layer, such a Janus membrane exhibits a higher atmospheric water capture capacity than those with homogeneous wettability, being a good candidate for alleviating water scarcity. Meanwhile, larger pores in the hydrophobic layer exhibit a higher atmospheric water capture capacity. Furthermore, the Janus membrane has been used in the process of plant cultivation. It is obvious that the Janus membrane is a good “water captor”, promoting plant growth. This work provides new insight into the novel design of the Janus membrane and broadening its real applications.

Author contributions

B. Ren, X. Zhang and R. Wang offered the idea and designed the experiments. B. Ren, H. Pi and J. Wu performed the experiments and analysed the experimental data. X. Zhao and M. Hu helped to analyse the data. J. Wu guided the study and revised the manuscript.

Conflicts of interest

The authors declare no conflict of interest.

Acknowledgements

The authors acknowledge the National Science Foundation of China (NSFC) (21503005), Beijing Municipal Natural Science Foundation (2154047), The Youth Outreach Project of Beijing (CIT&TCD201904058), and The Youth Outreach Project of Beijing Institute of Fashion Technology (BIFTBJ201806).

Notes and references

  1. M. M. Mekonnen and A. Y. Hoekstra, Science, 2016, 2, e1500323 Search PubMed.
  2. M. Verdaguer, M. Molinos-Senante, N. Clara, M. Santana, W. Gernjak and M. Poch, Sci. Total Environ., 2018, 624, 1308–1315 CrossRef CAS PubMed.
  3. H. Zhu, Z. G. Guo and W. M. Liu, Chem. Commun., 2016, 52, 3863–3879 RSC.
  4. H. Yang, J. R. Thompson and R. J. Flower, Nature, 2015, 525, 455 CrossRef CAS PubMed.
  5. N. Joshi, R. N. Dholakiya, M. A. Kumar and K. H. Mody, Environ. Prog. Sustainable Energy, 2017, 36, 1458–1465 CrossRef CAS.
  6. S. H. Zhou, F. Liu, J. Q. Wang, H. B. Lin, Q. Han, S. F. Zhao and C. Y. Tang, Environ. Sci. Technol. Lett., 2019, 6, 79–85 CrossRef CAS.
  7. S. H. Zhou, Z. Xiong, F. Liu, H. B. Lin, J. Q. Wang, T. T. Li, Q. Han and Q. L. Fang, J. Mater. Chem. A, 2019, 7, 632–638 RSC.
  8. Z. Q. Zhang, S. F. Li, B. X. Mi, J. B. Wang and J. N. Ding, Sci. Adv., 2020, 6, eaba9471 CrossRef CAS PubMed.
  9. H. Kim, S. Yang, S. R. Rao, S. Narayanan, E. A. Kapustin, H. Furukawa, A. S. Umans, O. M. Yaghi and E. N. Wang, Science, 2017, 358, 430–434 CrossRef PubMed.
  10. G. B. Hwang, A. Patir, K. Page, Y. Lu, E. Allan and I. P. Parkin, Nanoscale, 2017, 9, 7588–7594 RSC.
  11. A. R. Parker and C. R. Lawrence, Nature, 2001, 414, 33–34 CrossRef CAS PubMed.
  12. Y. F. Wang, X. W. Wang, C. L. Lai, H. W. Hu, B. Fei and J. H. Xin, ACS Appl. Mater. Interfaces, 2015, 8, 2950–2960 CrossRef PubMed.
  13. C. Dorrer and J. RüHe, Langmuir, 2008, 24, 6154–6158 CrossRef CAS PubMed.
  14. H. G. Andrews, E. A. Eccles, W. C. E. Schofield and J. P. S. Badyal, Langmuir, 2011, 27, 3798–3802 CrossRef CAS PubMed.
  15. Y. Chen and Y. M. Zheng, Nanoscale, 2014, 6, 7703–7714 RSC.
  16. T. Nørgaard and M. Dacke, Front. Zool., 2010, 7, 23 CrossRef PubMed.
  17. W. F. C. Blumer, R. Merlin and C. Gans, Amphibia-Reptilia, 1982, 3, 57–64 Search PubMed.
  18. P. Comanns, C. Effertz, F. Hischen, K. Staudt, W. Bhme and W. Baumgartner, Beilstein J. Nanotechnol., 2011, 2, 204–214 CrossRef CAS PubMed.
  19. Y. M. Zheng, H. Bai, Z. B. Huang, X. L. Tian, F. Q. Nie, Y. Zhao, J. Zhai and L. Jiang, Nature, 2010, 463, 640–643 CrossRef CAS PubMed.
  20. J. Ju, H. Bai, Y. M. Zheng, T. Y. Zhao, R. C. Fang and L. Jiang, Nat. Commun., 2012, 3, 1247–1252 CrossRef PubMed.
  21. K. Yin, H. F. Du, X. R. Dong, C. Wang, J.-A. Duan and J. He, Nanoscale, 2017, 9, 14620–14626 RSC.
  22. H. Zhou, M. X. Zhang, C. Li, C. L. Gao and Y. M. Zheng, Small, 2018, 14, 1801335 CrossRef PubMed.
  23. H. K. Raut, A. S. Ranganath, A. Baji and K. L. Wood, Appl. Surf. Sci., 2019, 465, 362–368 CrossRef CAS.
  24. S. C. Thickett, C. Neto and A. T. Harris, Adv. Mater., 2011, 23, 3718–3722 CrossRef CAS PubMed.
  25. M. Xiao, M. J. Cheng, Y. J. Zhang and F. Shi, Small, 2013, 9, 2509–2514 CrossRef CAS PubMed.
  26. L. Hu, S. J. Gao, Y. Z. Zhu, F. Zhang, L. Jiang and J. Jin, J. Mater. Chem. A, 2015, 3, 23477–23482 RSC.
  27. A. Khang, P. Ravishankar, A. Krishnaswamy, P. K. Anderson, S. G. Cone, Z. Liu, X. Qian and K. Balachandran, J. Biomed. Mater. Res., Part B, 2016, 105, 2455–2464 CrossRef PubMed.
  28. C. Shen, R. Wycisk and P. N. Pintauro, Energy Environ. Sci., 2017, 10, 1435–1442 RSC.
  29. H. X. Wang, J. Ding, L. M. Dai, X. G. Wang and T. Lin, J. Mater. Chem., 2010, 20, 7938–7940 RSC.
  30. J. Wu, N. Wang, L. Wang, H. Dong, Y. Zhao and L. Jiang, Soft Matter, 2012, 8, 5996–5999 RSC.
  31. H. Zhou, H. X. Wang, H. T. Niu and T. Lin, Sci. Rep., 2013, 3, 2964–2968 CrossRef PubMed.
  32. H. X. Wang, H. Zhou, H. T. Niu, J. Zhang, Y. Du and T. Lin, Adv. Mater. Interfaces, 2015, 2, 1400506 CrossRef.
  33. J. Wu, H. Zhou, H. X. Wang, H. Shao, G. L. Yan and T. Lin, Adv. Mater. Interfaces, 2019, 6, 1801529 CrossRef.
  34. Z. J. Wang, G. J. Liu and S. S. Huang, Angew. Chem., Int. Ed., 2016, 55, 1–5 CrossRef.
  35. Y. Liang, J. W. Shi, P. Xiao, J. He, F. Ni, J. W. Zhang, Y. J. Huang, C.-F. Huang and T. Chen, Chem. Commun., 2018, 54, 12804–12807 RSC.
  36. R. Z. Waldman, H. C. Yang, D. J. Mandia, P. F. Nealey, J. W. Elam and S. B. Darling, Adv. Mater. Interfaces, 2018, 5, 1800658 CrossRef.
  37. K. Yin, S. Yang, X. R. Dong, D. K. Chu, J.-A. Duan and J. He, ACS Appl. Mater. Interfaces, 2018, 10, 31433–31440 CrossRef CAS PubMed.
  38. F. Zhang, W. B. Zhang, Z. Shi, D. Wang, J. Jin and L. Jiang, ACS Appl. Mater. Interfaces, 2013, 25, 4192–4198 CAS.
  39. L. L. Hou, N. Wang, X. K. Man, Z. M. Cui, J. Wu, J. C. Liu, S. Li, Y. Gao, D. M. Li, L. Jiang and Y. Zhao, ACS Nano, 2019, 13, 4124–4132 CrossRef CAS PubMed.
  40. H. Fong, I. Chun and D. H. Reneker, Polymer, 1999, 40, 4585–4592 CrossRef CAS.
  41. J. Wu, N. Wang, L. Wang, H. Dong, Y. Zhao and L. Jiang, ACS Appl. Mater. Interfaces, 2012, 4, 3207–3212 CrossRef CAS PubMed.
  42. J. Wu, N. Wang, Y. Zhao and L. Jiang, Nanoscale, 2015, 7, 2625–2632 RSC.
  43. S. N. Tang, H. H. Pi, Y. Y. Zhang, J. Wu and X. Q. Zhang, Appl. Sci., 2019, 9, 3302–3314 CrossRef CAS.
  44. Z. X. Wang, X. B. Yang, Z. J. Cheng, Y. Y. Liu, L. Shao and L. Jiang, Mater. Horiz., 2017, 4, 701–708 RSC.
  45. X. L. Tian, H. Jin, J. Sainio, R. H. A. Ras and O. Ikkala, Adv. Funct. Mater., 2015, 24, 6023–6028 CrossRef.
  46. H. Zhou and Z. G. Guo, J. Mater. Chem. A, 2019, 7, 12921–12950 RSC.
  47. L. L. Yan, X. B. Yang, J. Long, X. Q. Cheng, D. Pan, Y. F. Huang and L. Shao, Chem. Commun., 2020, 56, 478–481 RSC.
  48. H. Bai, L. Wang, J. Ju, R. Z. Sun, Y. M. Zheng and L. Jiang, Adv. Mater., 2014, 26, 5025–5030 CrossRef CAS PubMed.
  49. Y. Y. Song, Y. Liu, H. B. Jiang, S. Y. Li, C. Kaya, S. Thomas, Z. W. Han and L. Q. Ren, Nanoscale, 2018, 10, 3813–3822 RSC.
  50. R. Rioboo, M. Marengo and C. Tropea, Exp. Fluids, 2002, 33, 112–124 CrossRef.

Footnote

Electronic supplementary information (ESI) available: Fig. S1–S4, S6, S7 and Tables S1–S3. See DOI: 10.1039/d1nr01120k

This journal is © The Royal Society of Chemistry 2021