Stimuli-responsive surfaces for switchable wettability and adhesion

Abstract

Diverse unique surfaces exist in nature, e.g. lotus leaf, rose petal and rice leaf. They show similar contact angles but different adhesion properties. According to the different wettability and adhesion characteristics, this review reclassifies different contact states of droplets on surfaces. Inspired by the biological surfaces, smart artificial surfaces have been developed which respond to external stimuli and consequently switch between different states. Responsive surfaces driven by various stimuli, e.g. stretching, magnetic, photo, electric, temperature, humidity and pH, are discussed. Studies reporting on either atmospheric or underwater environments are discussed. The application of tailoring surface wettability and adhesion includes microfluidics/droplet manipulation, liquid transport and harvesting, water energy harvesting and flexible smart devices. Particular attention is placed on the horizontal comparison of smart surfaces with the same stimuli. Finally, the current challenges and future prospects in this field are also identified.

1. Introduction

Wettability is an important property of solid surfaces, which has a profound impact on the fields of self-cleaning [1,2], microfluidic channels [3], anti-fouling [4], fog harvesting [5], anti-icing [6], oil/water separation [7], water energy harvesting [8] and thermal management [9]. In nature, some surfaces show similar contact angles but possess different levels of adhesion. For example, the upper side of a lotus leaf displays superhydrophobic low adhesion surfaces, which are slippery where water droplets can freely move [10,11]. However, the surfaces of rose petals show high adhesion, on which the water droplets are pinned, though the surface is superhydrophobic [12]. Rice leaves possess superhydrophobic anisotropic adhesion surfaces, on which droplets are influenced by different forces when moving in different directions [13]. As for some aquatic organisms, for example, filefish [14] and clam shell [15] have underwater superoleophobic surfaces with low oil adhesion, where oils can easily leave the surface.

Extensive research has been reported in the literature investigating how diverse micro/nano-structures affect surface wettability, adhesion and liquid behaviours on the surfaces [16–18]. In the past decade, researchers have also fabricated a series of liquid-infused surfaces by injecting lubricant into porous, rough or swellable substrates [6]. To develop multifunctional and intelligent surfaces, there have been many studies concentrating on the constructing of stimuli-responsive materials, where wettability and adhesion can be switched reversibly are focused over the past several years (figure 1).

Figure 1.

Principle of stimuli-responsive surfaces: stimuli factor (mechanical stretching, magnetic, photo, electric, temperature, etc.) result in change in micro/nano-structure, chemical composition or chemical distribution; consequently, surface adhesion (contact state) is changed.

In this literature review, concepts related to surface wetting for both the atmospheric and underwater environments are included. The relationship between wettability, adhesion and liquid behaviour is summarized. As shown in figure 1, surfaces with diverse wettability and adhesion are reclassified into four contact states. Particular attention is paid to responsive switchable-adhesion surfaces, including surfaces that respond to the stimuli of stretching, magnetism, light, electricity and temperature. By reviewing the studies on the responsive surfaces, the potential for material design is summarized and suggestions for future studies are recommended.

2. Theory basis

2.1. Wettability

Wettability is the property of a surface that describes the degree to which a surface can be wet by a liquid. This property can be quantified by the contact angle (CA), i.e. the value measured through the liquid where an environmental phase–liquid interface meets a solid surface. Generally, a solid surface with a CA < 90° is considered to be lyophilic, and a solid surface with a CA > 90° is deemed to be lyophobic. For example, when the environment is atmospheric and the liquid is water, a surface with a water contact angle (WCA) > 90° is called a hydrophobic surface. The critical angle of 90° between lyophilic and lyophobic is derived from Young's equation [19]. However, Yoon et al. [20] indicated that the critical angle of hydrophilicity and hydrophobicity was around 65° through the testing of local changes in the chemical potential of water using assistive techniques such as a surface tension meter. Jiang et al. [21] further deduced that the critical limit CA of hydrophilicity and hydrophobicity should be around 62.7°.

In addition, a surface with CA < 5° is defined as superlyophilic, and a solid surface with CA > 150° is defined as superlyophobic. For example, when the environment is underwater and the liquid is an oil, a surface with oil CA (OCA) > 150° is called an underwater superoleophobic surface.

2.2. Adhesion and contact state

Adhesion, in a broad sense, can refer to the properties of an adhesive joint (solid–adhesive–solid system), or to a combination via van der Waals force (dry adhesion), surface absorption and attraction [22–25]. Herein, we focus on any surface-induced obstacles that hinder liquid flow on the surface in any direction or away from the surface, quantified by CA hysteresis (CAH), sliding/rolling angle (SA) or relevant force named as adhesive force, retention, resistance or friction [26,27].

Surface adhesion is closely related to wettability: on a uniform smooth surface, the larger the contact angle, the smaller the adhesion. However, the situation on practical surfaces is complicated because the micro/nano-structure on the surface may affect the formation of the three-phase contact line (TPCL). A typical example is a rose petal on which the water droplets are pinned, though the surface is superhydrophobic [12]. In fact, both adhesion and wettability of the surface are firmly associated with surface chemical composition, chemical distribution and topography. By changing these factors, surface adhesion to liquid can be adjusted to control the contact situation between the liquid and the surface. Based on the difference in wettability and adhesion, here we divide the liquid behaviour on a surface into four typical contact states (table 1 and figure 1).

Table 1.

Different surface contact states and the corresponding liquid behaviour.

contact state	description of the state	liquid behaviour on surface
lyophilic	ultra high adhesion	droplet can wet the surface or even spread flat
pinning (rose petal)	lyophobic but high adhesion	droplet turns spherical, and stays still on the surface
slippery (lotus leaf)	lyophobic and low adhesion	droplet turns spherical, and can easily slide/roll off
anisotropic (rice leaf)	high retention for one direction and low retention for another	droplet tends to move directionally

3. Responsive surfaces

Artificial smart surfaces, which are designed to possess stimuli-responsive wettability and adhesion, and can transfer from one state to another are introduced in this section.

3.1. Stretching-responsive surfaces

Mechanical stretching (strain) can alter the bulk and surface geometry of a soft material, consequently enabling the reversible control of surface adhesion. Simply enlarging or compressing the material will not change the surface tension or chemical composition [28]. However, the surface roughness will be notably changed, which would influence the CAH and other adhesion properties of the surface, resulting in different liquid behaviour (table 2) [33–35].

Table 2.

Comparison of typical stretching-responsive surfaces.

structural feature	advantages	ref.
responsive SLIPS	can realize SA change without CA change, switching between pinning and slippery state	[29]
micropillar arrays structured soft substrate	can realize CAH change and droplet deformation	[30]
asymmetric stretching on soft materials	switching between anisotropic and pinning state, can realize unidirectional droplet motion	[31]
soft striped-microgroove surface	can realize SA change without CA change, switching between pinning and slippery state, can realize unidirectional droplet motion	[32]

The strain response can be used on a slippery lubricant-infused porous surface (SLIPS) [36–39], where either pinning or the slippery state can be obtained, respectively. Yao et al. combined a porous Teflon nanofibre network with an elastic polydimethylsiloxane (PDMS) membrane to obtain a stretchable substrate [29]. Perfluoropolyether was then infused into the porous substrate as the lubricant so that the stretchable SLIPS was achieved. When the SLIPS was compressed, the diminished nanopores contributed to lubricant release. The lubricating layer (figure 2a,b) gave rise to a much lower adhesion surface (slippery state). By elongating the SLIPS via external mechanical tension, the nanopores were enlarged. Lubricant flowed to the nanopores, resulting in a higher adhesion surface causing water pinning (figure 2b,c). As shown in figure 2d, the SA of different liquid on such a mechanical stretching response surface could range from 5° to 25°. This kind of material has potential for application in the field of droplet manipulation.

Figure 2.

Mechanical stretching response on smart materials (in air). (a–d) Stretching response on a SLIPS, from slippery to non-slippery. Adapted from [29]. Copyright © 2013, Springer Nature. (e–g) Stretching response on a micropillar array surface, switch in CA and CAH. Adapted from [30]. Copyright © 2017, American Institute of Physics. (h,i) Asymmetric-stretching response on a stripe-microgroove surface, unidirectional adhesion. Adapted from [31]. Copyright © 2014, Wiley-VCH. (j) Stretching response on a soft skin-like surface; switch between lotus leaf state and rose petal state. Adapted from [32]. Copyright © 2018, Wiley-VCH.

Coup et al. [30] studied the stretching response on a micropillar arrays structured soft PDMS substrate. They found that surface wettability and CAH changed as a function of surface dilation (Σ). When the substrate was elongated, the shape of the micropillars was slightly changed but the intensity of micropillars was sharply reduced (figure 2e). As shown in figure 2f,g, such an intensity reduction finally caused a lower contact angle and a higher surface adhesion. Su et al. [40] fabricated a Janus PDMS and found that the Laplace driving force and wettability-gradient driving force for water movement could both be controlled through stretching or compressing the substrate. It was suggested that such a material will find potential applications not only in fog harvesting, but also in dynamic fog-flux regulators.

The strain response can also be designed on soft anisotropic materials. Zhang et al. [31] studied the asymmetric-stretching response on a stripe-microgroove structured organogel composed of poly(butyl methacrylatecolauryl methacrylate) and silicone oil. Unidirectional droplet motion was achieved via asymmetrically stretching the soft organogel. As shown in figure 2h, when the organogel was elongated vertical to the groove direction from one side, the water droplet can easily move (water slide angle WSA ≤ 10°) from the elongated side to non-elongated side. By contrast, the water droplet was pinned when it comes to moving against (figure 2i). They suggested that the switch in adhesion was caused by the asymmetric-stretching induced fan-shaped grooves, which act as a physical barrier that guides droplet movement. The roughness variation had only a minor effect. Wang et al. [32] studied the bending and stretching response on a soft skin-like (striped microgroove) PDMS. When the strain was applied perpendicular to the stripe direction, WCA remained over 140° but WSA ranged from 3° to the pinning state, as shown in figure 2j. The surface was reversely switched from the lotus leaf state to the rose petal state. Taking advantage of these soft materials, the stretching-responsive surfaces can be further designed as wearable devices, or as other flexible smart devices.

3.2. Magnetic-responsive surfaces

3.2.1. Atmospheric environment

A magnetic substance can be incorporated into the substrate during manufacture to obtain a magnetic-responsive surface. The surface roughness and/or microstructure could then become controllable via an external magnetic field, so as to adjusting the wettability and adhesion (table 3) [47–49]. By controlling these factors, droplet manipulation, water collection or liquid transport can be realized. Also, these controllable surfaces would be useful for designing or manufacturing microfluidic devices.

Table 3.

Comparison of typical magnetic-responsive surfaces.

structural feature	control avenue	advantages	ref.
magnetic fluid in nanoarray (in air)	adjusting the intensity of magnetic field	can control CA, realizing unidirectional droplet motion, antigravity motion and motion at different speeds	[41]
flexible magnetized conical array (in air)	adjusting the direction of magnetic field	can realize change in SA and retention force, can realize and control droplet directional bouncing	[42,43]
FLIPS (in air)	applying or removing magnetic field	can realize SA change to control droplet displacement, switching between pinning and slippery state	[44,45]
two uniform-nanorods structured substrates (underwater)	applying periodic magnetic field	can realize reversibly oil droplet transferring between substrates	[46]

Generally, to make a material magnetic-responsive, a magnetic substance needs to be added. Tian et al. [41] added an Fe₃O₄ magnetic fluid into a ZnO nanoarray to obtain a magnetic-responsive surface. As shown in figure 3a, by adjusting the intensity of the external magnetic field, the surface could be switched between having a smooth and rough microstructure, showing different contact angles and adhesion. By further varying the magnetic field, unidirectional droplet motion, antigravity motion, and motion at different speeds were easily realized.

Figure 3.

Magnetic response on smart materials (in air). (a) Response on magnetic fluid/ZnO nanoarray composite interface, switch in surface roughness, CA and adhesion. Adapted from [41]. Copyright © 2016, American Chemical Society. (b) Response on conical array structured surface, switch in microstructure. Adapted from [50]. Copyright © 2015, Wiley-VCH. (c–f) Response on FILPS, switch in macro/microstructure, adhesion force and adhesion performance. Adapted from [44]. Copyright © 2018, Springer Nature. (g–j) Response on lubricant-containing conical array structured surface, switch between pinning state and slippery state. Adapted from [45]. Copyright © 2017, Wiley-VCH.

A conical array structured surface can be magnetized to become a smart material. Peng et al. [50] added Co magnetic particles to PDMS during the curing process of the soft lithography to obtain a magnetic-responsive microcone-array structured surface. As shown in figure 3b, the conical arrays were vertical when the external magnet was placed directly above the surface. Moving the magnet to the right or left, the conical arrays were attracted to the magnet to the same direction. Also, by applying the magnetism as a cyclic function, periodic vibration of the microcone arrays was achieved, improving fog capturing performance due to the optimized contact area with moisture and the Laplace driving force. If required, the permanent bending of the micro-arrays could also be realized using this method. Li et al. [42] made use of the magnetic response of Co particles during the preparation and achieved arrays which displayed diverse permanent obliquity. The WSA ranged from over 60° to below 5° and the retention force ranged from over 80 µN to below 20 µN on account of the different obliquity. They also found that a water droplet showed a directional bouncing trend on the bent arrays. The horizontal displacement sharply increased, and the bounce height slightly decreased with the arrays bending from vertical to 45° obliquity. Lin et al. [43] realized diverse permanent bending by means of magnetization and indicated that a very slight vertical vibration (e.g. wind, sound) would cause droplet movement towards the bending direction.

A magnetic response could also be imparted to SILPS by using magnetic substances as lubricants. Wang et al. [44] employed a ferrofluid when preparing SILPS (namely FLIPS). When no magnetic field was applied, magnetic nanoparticles were aligned randomly, showing a relatively smooth surface (figure 3c). However, applying an external magnetic field gradient could preferentially align the magnetic nanoparticles in the ferrofluid. As shown in figure 3c,d, the ferrofluid showed a multiscale hierarchical topographical shape. Consequently, the adhesion force (figure 3e) against water movement was controlled accurately and reversibly. For example, figure 3f shows the totally different adhesion properties to droplets when diluted ferrofluids are employed.

Huang et al. [45] combined the concept of FILPS with the magnetic-responsive microcone-array (figure 3g). As shown in figure 3h, the micro-array could be deformed between the states of being flat (0°) and perpendicular (90°) by application of the external magnetic field. The droplet bouncing and sliding test showed that, when the micro-array was perpendicular to the surface, the droplet was pinned to the surface (figure 3i), indicating high adhesion. Adjusting the magnetic field to make the micro-array lie down, the droplet came into nearly full contact with the slippery lubricant, releasing the droplet (figure 3j). Li et al. [51] used a similar strategy and produced a PDMS@Fe₃O₄-coated fabric. They found that different Fe₃O₄ concentrations could lead to different values of WCA. By adjusting the magnetic response, the surface wettability and adhesion could be reversibly transformed.

3.2.2. Underwater environment

Recently, Feng et al. [46] demonstrated a method for driving oil-based magnetic fluids using an external magnetic field. They first used a low-temperature hydrothermal method to grow a uniform array of ZnO nanorods on a glass substrate. This ZnO surface showed superoleophobicity underwater after UV treatment (OCA = 156°, oil slide angle OSA < 5°, figure 4a). Then, they used two ZnO surfaces with the above properties to move oil droplets in water. As shown in figure 4b, one substrate (top) was suspended above the other substrate (bottom), making them opposite and leaving space between them. The magnetic oil droplets (i.e. silicone oil containing Fe₃O₄ nanoparticles) were injected onto the lower substrate. By applying a periodic external magnetic field to the system as a driving force, they found that the oil droplets were able to quickly respond to the stimulation of the external magnetic field, reversibly transferring between the two substrates. During the entire oil droplet movement, there was no adhesion of magnetic oil droplets to the substrate due to the low oil adhesion of the ZnO nanorod array to the magnetic oil. Conversely, when two smooth ZnO substrates were used for the same transmission experiment, a noticeable delay in the movement of the oil droplets was observed. These findings might be useful for oil/water separation and for underwater anti-fouling devices.

Figure 4.

Magnetic response on smart surfaces (underwater). (a) Morphology and wettability characterization of the aligned ZnO nanorod arrays. (b) The manipulation and transfer of an underwater oil droplet by a magnetic fluid droplet on the aligned ZnO nanorod arrays. Adapted from [46]. Copyright © 2016, The Royal Society.

3.3. Temperature-responsive surfaces

3.3.1. Atmospheric environment

Increasing the temperature may give rise to the partial melting of a solid which then introduces a liquid layer between the solid surface and the droplet so that the surface adhesion switches [15,52]. Yao et al. [53] realized a temperature-responsive surface adhesion on a n-paraffin-swollen organogel surface. At room temperature, the rough surface was in a high adhesion state with high adhesive force (over 200 µN) and water droplet pinning. When the temperature was raised to 58–60°C, the n-paraffin partly transformed to the liquid phase, making the surface slippery. The surface transformed into a low adhesion state with a low adhesive force (around 40 µN) and a WSA of approximately 6°. Decreasing the temperature could then solidify the n-paraffin to realize the reversible switch of surface adhesion.

Increasing the temperature to the glass transition temperature of a flexible polymer can activate molecular chain mobility and alter the elasticity of the polymer [54–58]. Employing such a flexible polymer with shape memory can achieve the reversible control of surface microstructure, which further controls the adhesion [54,59,60]. For instance, Cheng et al. [61] fabricated a micro-array structured surface using a shape memory polymer. As shown in figure 5a, the micro-array was uniform and vertical, and the surface showed a superhydrophobic (WCA = 151° ± 1.5°) isotropic adhesion (WSA = 24° ± 1.2°) similar to the lotus leaf. When the surface was thermally pressed at a temperature slightly above the glass transition temperature, the pressed pillars laid down, turning into a microgroove-like structure. The surface showed an anisotropic adhesion property similar to the rice leaf. The WSA in the A direction was 20° ± 1.3°, and the WSA in the B direction was 29° ± 1.5° (figure 5a). When the surface was reheated under the same temperature, the pillar stands vertically, thus the reversible switch between superhydrophobic isotropic adhesion and anisotropic adhesion is achieved (figure 5b). They also realized the laying down of selected pillars on the surface, allowing for control of water movement. Such smart surfaces may potentially have application in bio-separation or in microfluidic devices.

Figure 5.

Temperature/thermal response on smart surfaces (in air). (a,b) On a micropillar array structured shape memory polymer, adhesion switch between isotropy to anisotropy. Adapted from [61]. Copyright © 2018, Wiley-VCH. (c,d) On a bioinspired peristome-beak structured PINPAAM based microgel, switch in adhesion force value and direction giving rise to change in water spreading behaviour. Adapted from [62]. Copyright © 2018, Wiley-VCH.

Increasing the temperature to the lower critical solution temperature (LCST) for some block copolymers can give rise to a polymeric transformation from the extended state to the collapsed state [63–65]. N-isopropylacrylamide (PNIPAAm)-based polymer is a good example, where the surface roughness may change with increasing temperature. More significantly, PNIPAAm-based material will transform the hydrogen bonding between intra- and intermolecular under different temperatures [66,67], causing a notable change in wettability and adhesion. Hou et al. [68] found that the values of WCA and CAH on PMMA-b-PNIPAAm (PMMA: poly(methyl methacrylate)) were 51.64° and 43° at 20°C, respectively. When the temperature was raised to 45°C, the surface morphology of the polymer changed as the polymeric state transformed. The values of WCA and CAH increased to 107.09° and 60°, respectively, indicating a correlation between the temperature and the surface wetting property. Li et al. [62] fabricated a peristome-beak surface inspired by the nepenthes and shorebird using PNIPAAm-based microgel. Water showed different spreading behaviour on the surface due to the temperature-responsive switch in adhesion force value and direction. As shown in figure 5c, when the temperature was 20°C, the droplet unidirectionally flowed along the striped microgrooves. As shown in figure 5d, when the substrate was heated to 40°C, the droplet spread bidirectionally across the microgrooves, and was pinned at the overhangs. Such designs may find application in not only microfluidic but also in heat-transfer systems.

Temperature changes may also have complex effects upon the results of wettability and adhesion tests, especially in an atmospheric environment (table 4). It is noteworthy that on an overheated surface, water droplets may be levitated by their self-generated vapour. This phenomenon is called the Leidenfrost effect [74]. Zhong et al. [72] reviewed how surface topography and wettability affected the Leidenfrost effect, with a view to droplet geometry and dynamics. Recently, Graeber et al. [73] observed an even more complex motion of droplets on an overheated surface and referred to this as Leidenfrost droplet trampolining. However, heating a substrate sometimes causes notable variation in environmental humidity and the liquid physico-chemical properties [75–78]. These effects should not be ignored, and therefore should be investigated in future research. In addition, the temperature-related response could also be designed as an electrothermal response and a photothermal response. These are discussed in the following two sub-sections.

Table 4.

Comparison of typical temperature-responsive surfaces.

control avenue	applicable materials	advantages	ref.
partial melting of a solid surface introducing a liquid layer/solidifying the layer (in air)	organogel, e.g. n-paraffin-swollen organogel	can realize change in SA and adhesive force	[15,52,53]
increasing/decreasing the temperature to the glass transition temperature (in air)	thermoelastic polymer (microstructured)	switching between isotropic and anisotropic state, can realize droplet directional motion	[54,59–61]
increasing/decrease the temperature to the LCST (either in air or underwater)	block polymers, e.g. PNIPAAm-based polymer	can realize change in CA/adhesion, can control the direction of droplet motion	[62,68–71]
Leidenfrost effect (in air)	heat-resisting materials	can control droplet geometry and dynamic	[72,73]

3.3.2. Underwater environment

In addition to the research carried out in air, PNIPAAm-based materials are also employed for underwater studies. Heating the material to the LCST, the adhesion of oil droplets to the surface will be changed. Oil droplets will then show different behaviour on the surface. Therefore, underwater oil droplet manipulation can be realized, which can be further applied to oil/water separation and underwater anti-fouling surfaces.

Chen et al. [71] prepared a thermally responsive PNIPAAm hydrogel. The surface wettability and the adhesion to oil droplets were related to the temperature of the surrounding aqueous solution. When the temperature of the solution was lower than the LCST (32°C) of PNIPAAm, the surface showed underwater superoleophobicity and low oil adhesion, i.e. the underwater OCA was 151.7° ± 1.6°, and oil adhesion force was 5.8 µN. When the temperature of the solution was raised to above the LCST, the surface showed weak oleophobicity and high oil adhesion. At this time, the underwater OCA of the hydrogel decreased to 127.0° ± 4.6°, and the adhesion increased to 23.1 µN. They also found that the two different states could be switched reversibly multiple times. The PNIPAAm hydrogel displayed this performance due to the large number of hydrophobic groups (–CH(CH₃)₂) and hydrophilic groups (–CONH–) contained in its network. At room temperature (<LCST), the N–H/C=O groups on the branch of PNIPAAm formed an intermolecular hydrogen bond with water molecules in the solution, which made the gel hydrate and swell. The trapped water layer further realized the superoleophobicity and low oil adhesion. However, when the solution temperature was raised above the LCST, the N–H/C=O groups in the hydrogel formed intramolecular hydrogen bonds with the molecules on the adjacent branches. This formation made the hydrogel dehydrate and collapse, losing the repulsion to oil droplets, finally reducing the oleophobicity and enhancing the adhesion to oil.

Liu et al. [70] reported the silicon nanowire array-modified heptafluorodecyltrimethoxysilane (HFMS) and heat-sensitive PNIPAAm (figure 6a). The switch between superoleophilic state and superoleophobic state of the surface was achieved by adjusting the water temperature. As shown in figure 6b, the surface was superoleophobic in water at 20°C, and the underwater OCA (1,2-dichloroethane, 2 µl) was 157°. However, when the ambient temperature increased, it became superoleophilic, and the underwater OCA was only 3° at 60°C. They also found that the transition between the underwater superoleophobicity (20°C) and the underwater superoleophilicity (60°C) was reproducible (figure 6c). The switchable-adhesion is mainly derived from the grafted PNIPAAm chain which could cover or expose oleophilic HFMS components through the change of thermally responsive molecular structure. At room temperature, the PNIPAAm chain preferentially bonded with water molecules in solution to form intermolecular hydrogen bonds. The hydrated PNIPAAm chain tended to extend further outward, so that the oleophilic HFMS chain was covered and hidden, causing underwater superoleophobicity. As the temperature was increased to 60°C, the PNIPAAm chain mainly generated intramolecular hydrogen bonds among its branches, which caused the dehydration shrinkage of the grafted PNIPAAm chain, exposing the oleophilic HFMS chain to form an underwater superoleophilic surface (figure 6c,d).

Figure 6.

Temperature response on smart surfaces (underwater). (a–d) PNIPAAm-HFMS surface. Adapted from [70]. Copyright © 2015, Wiley-VCH. (a,b) Underwater temperature-responsive PNIPAAm-HFMS surface, and the changes of underwater contact angles are repeatable. (c) Real-time underwater contact angles of DCE droplets with different chemical compositions at 60°C. (d) Underwater temperature-responsive switch between low and high adhesion on PNIPAAm-SiNWA. (e–g) PNIPAAm hydrogel. Adapted from [69]. Copyright © 2019, Wiley-VCH. (e) The conformational change mechanism of the PNIPAAm polymer chains with water molecules responsible for the thermo-responsive curving behaviour of the hydrogel film. (f) Schematic diagram and snapshot showing dynamic sliding of oil drop on fibrillar side and non-structured side of the hydrogel film at high (40°C) and low (9°C) temperature condition. (g) Reversibly capturing and releasing the oil droplet by thermo-responsive bending of the hydrogel film underwater.

Inspired by the predation mechanism of sundews in nature, Ma et al. [69] constructed a new PNIPAAm-based hydrogel actuator with high mechanical strength and asymmetric structure (figure 6e). The hydrogel actuator could not only respond to environmental temperature changes (LCST of PNIPAAm) and achieve reversible switching of surface adhesion (figure 6f), but also could realize fast solvent exchange between ethanol and water. In addition, the actuator was able to perform reversible responsive bending, accompanied by an actuating force (≈21 mN), which enabled the actuator to reversibly capture and release underwater oil droplets (figure 6g).

3.4. Electric responsive surfaces

3.4.1. Atmospheric environment

An electric field can affect surface wettability, adhesion and liquid behaviour in various ways (table 5). For example, the WCA can be reduced by applying an electric field between a conducting droplet and a counter electrode underneath the droplet, which is an electrowetting course without changing the surface structure [28,79]. Previous studies indicated that by changing the applied voltage during electrowetting, the CA could be accurately adjusted; the adhesion and liquid behaviour could also be controlled using this strategy [80].

Table 5.

Comparison of typical electric responsive surfaces (in air and underwater).

control avenue	materials	surface topography	characteristic after electric field applied	advantages	ref.
electrowetting effect	conductor	no change	CA reduces	can accurately adjust CA and control droplet motion	[28,79,80]
	SLIPS using conducting substrates	no change	pinning state (effectiveness of lubricants reduces)	can realize the switch between pinning state and slippery state	[81,82]
	conductive polypyrrole (PPy) film	no change	underwater OCA increases	can control the movement of underwater oil droplets or bubbles	[83]
new electro-dewetting mechanism	superhydrophilic conductor	no change	CA increases	can realize droplet directional moving, merging and splitting	[84]
effect of electric potential on dodecyl chains	dodecyl chain-based materials	change	superhydrophilic state (electric potential rotated hydrophobic dodecyl)	can realize the switch between superhydrophilic state and slippery state	[85]
electric-thermal expansion effect	SLIPS using flexible substrates	change	pinning state (lubricants penetrate into expanded porous)	can further realize complex droplet behaviour, e.g. oscillation, jetting	[86,87]
electrochemical reaction	binary composite metal surface	change	underwater OCA reduces	can control the oil spreading on surface	[88]
electrochemical oxidation	conductive polymer PANI	no change	underwater OCA reduces	can control the oil adhesion on surface	[89]

Guo et al. [81] applied a bar electrode above a SLIPS where the upper side of the droplet was in contact with the electrode. When there was no applied voltage, the surface was in the slippery state due to the lubricant (figure 7a). When the voltage was applied, the microstructure of the surface was not changed but the droplet remained on the surface (figure 7b). They suggested that the switch in adhesion was caused by the electrostatic attraction between the SLIPS and the liquid, which reduced the effectiveness of the lubricant. Wang et al. [82] further found that the responsive voltage could be reduced by employing lubricants with a lower viscosity during SLIPS fabrication.

Figure 7.

Electric response and the wetting on the surfaces (in air). (a,b) Electrowetting. Adapted from [81]. Copyright © 2016, Wiley-VCH. (c) Electric corona discharge controlled liquid behaviour. Adapted from [90]. Copyright © 2019, American Institute of Physics. (d) On Cu electrode-embedded substrate, electric responsive fast moving of large ink drop. Adapted from [91]. Copyright © 2019, American Chemical Society. (e,f) Electro-dewetting and its application on droplet directional moving, merging and splitting. Adapted from [84]. Copyright © 2019, Springer Nature. (g) On a porous membrane with dodecyl chains, switch between superhydrophobic low adhesion and superhydrophilic. Adapted from [85]. Copyright © 2019, American Association for the Advancement of Science. (h) On a SILPS using shape memory graphene film as the substrate, switch between slippery state to non-slippery state. Adapted from [87]. Copyright © 2017, American Association for the Advancement of Science. (i–k) On a sandwiched substrate, adhesion switch, droplet oscillation and jetting. Adapted from [86]. Copyright © 2018, Wiley-VCH.

Li et al. [90] controlled liquid motion via an electric corona discharge on a polyethylene terephthalate (PET) film-coated Cu electrode. They drilled a hole into the PET where the conductive substrate was exposed to the direct current (DC) power needle. A silicone oil droplet was released onto the PET surface near the hole and it showed a flat-ellipsoid shape. When the DC power was applied, the droplet immediately jumped into a spherical shape, moving horizontally away from the hole, as shown in figure 7c. The displacement of the oil droplet was affected by the distance between the droplet and the conductive hole. Furthermore, Plog et al. [91] studied the electric responsive movement of ink droplets which have diverse chemical composition and droplet size on the Cu electrode-embedded substrate (figure 7d). They realized the accurate motion control of commercially pure-liquid, carbon fibre suspension and polymer solution ink droplet in 200 µm–3 mm diameters by adjusting the applied voltage.

Recently, Li et al. [84] developed a novel mechanism termed electro-dewetting. They employed a bare silicon wafer as a highly hydrophilic substrate and an ionic surfactant as a droplet. In stark contrast to conventional electrowetting, the contact angle did not decrease but rather increased when the voltage and current were applied. It was explained that the interaction between the droplet and the conductive substrate was not directly controlled by the voltage. As shown in figure 7e, the electro-dewetting probably resulted from the electric field-induced adhesive interaction change of the ionic surfactants to the hydrophilic substrate. By changing the direction of the electric field, the ionic surfactant molecules migrated towards or away from the hydrophilic substrate, realizing electro-dewetting and rewetting. Under this novel mechanism of electric field responsive change of adhesion, droplet directional moving, merging and splitting were realized, as shown in figure 7f. This finding opens the door to a wide range of potential applications related to microfluidics.

An electric field can also adjust the surface wettability and adhesion by switching the structure in some smart materials [92,93]. Liu et al. [85] used the effect of electrical potential on the dodecyl chains to fabricate an electric responsive porous membrane. The electric potential rotated the hydrophobic dodecyl chains to conceal the pores. This transformation enables the switch in surface wettability from superhydrophobic low adhesion (WCA around 152°, WSA around 4°) to superhydrophilic state (WCA near 0°). They also found that an organic reagent rinsing process could re-rotate the hydrophobic dodecyl chains to recover the superhydrophobicity, realizing the reversible switch in surface property (figure 7g). They found that such a surface could act as a smart liquid gate for sensors and find application for the efficient water/oil separation, and for controllable liquid transfer. An electric response can also be obtained by electric-thermal expansion. For example, Wang et al. [87] prepared a SILPS using shape memory graphene (GO) film as the substrate. The GO could be compressed to a desired initial length. Then, a constant DC voltage was applied, causing an electric-thermal effect on the GO film, which gave rise to the expansion of the compressed film. When the film was expanded, lubricant penetrated into the porous structure so that the surface switched from slippery (WSA around 2°) to non-slippery state (pining on the surface). The expansion of the GO film could be controlled by the applied voltage so that the surface adhesion was also tunable (figure 7h). Oh et al. [86] fabricated a SLIPS using a sandwich (i.e. an elastomeric membrane between two elastic electrodes) structured substrate which exhibited a dynamic variation in shape under varying voltage. The surface was in the slippery state when there was no voltage (figure 7i). Applying the voltage gave rise to the in-plane stretching of substrate. The rough structure without a covering of lubricant then appeared, causing the droplet to be pinned (figure 7i). Taking advantage of such an electric responsive switch in surface adhesion, more complicated droplets manipulation was realized, such as oscillation (figure 7j) and jetting (figure 7k).

3.4.2. Underwater environment

Underwater electro-responsive surfaces are useful for manipulating liquids/bubbles in various applications, such as microfluidics, oil recovery, printing/patterning drop control, underwater anti-fouling and self-cleaning.

Wang et al. [88] proposed a strategy to achieve an in situ reversible adhesion transition between the underwater superoleophilic and superoleophobic states by constructing a binary textured surface. Taking the Cu/Sn composite system as an example, when closing and opening an external circuit, the reversible transition was realized (figure 8a). The underlying principle was that the electrodeposited Sn layer caused a water-deficient layer by suppressing the hydrogen bonding network at the water/electrode interface, resulting in the diffusion of oil droplets on the surface. When the potential was removed, Sn was converted into Sn²⁺ and dissolved in the electrolyte solution again, and the composite surface system would gradually restore to a Cu-rich surface with high affinity for hydroxyl groups, which replaced the oil layer with aqueous electrolyte solution (figure 8b). This material provides a new pathway for designing smart materials in the fields of liquid manipulating.

Figure 8.

Electric response on smart surfaces (underwater). (a,b) Cu/Sn composite surface. Adapted from [88]. Copyright © 2020, Springer Nature. (a) In situ reversible superwettability transition between underwater superoleophilicity and superoleophobicity. (b) The schematics show that the Cu/Sn binary composite surface enables drastic altering of the hydrogen bonding network at the electrolyte/electrode interface, making the oil droplet experience the in situ reversible transition between superoleophobicity and superoleophilicity. (c–e) Porous PS film. Adapted from [94]. Copyright © 2016, Wiley-VCH. (c) Morphology characterization of the gradient structured porous polystyrene (PS) film. (d) CA change for water and underwater oil droplets as functions of the applied voltage on the porous PS film surface. (e) Electric field and gradient microstructure cooperate to drive directional motion of an underwater oil droplet. (f–i) Porous PS films on copper wire. Adapted from [83]. Copyright © 2018, Wiley-VCH. (f) Morphology characterization. (g,h) Adhesion characterization. (i) Schematics of the wetting states of an underwater oil droplet (liquid paraffin) or air bubble on the porous PS film-coated copper wire surface under an applied electric field.

Polyaniline (PANI) is a conductive polymer, and there are some stable oxidation forms of it: semi-oxidized emeraldine base (EB) which could convert into emeraldine salt (ES) through a portion, fully oxidized pernigraniline base (PNB) and fully reduced leucoemeraldine base (LEB) [89,95–100]. Such a polymer could be employed for fabricating a responsive surface. For example, Ding et al. [89] prepared uniformly distributed PANI nanowire thin films (141.3 ± 5.0 nm in length and 45.7 ± 1.7 nm in diameter) through the aligned current deposition process. When a voltage of 0.43 V was applied to the PANI nanowire film, the OCA of 1,2-dichloroethane droplets in 0.1 mol l⁻¹ HClO₄ solution (pH = 1.2) on the surface is 161.61° ± 1.5° and the corresponding OSA was less than 3°, exhibiting superoleophobicity and a low oil adhesion in solution. When the voltage was adjusted to −0.2 V, although the surface still showed underwater superoleophobicity, its underwater OCA decreased to 154.2° ± 2.9°. No matter how the surface was inclined, the oil droplets are tightly adhered to the surface without slipping, indicating that the surface shows a high oil adhesion state in the solution under this condition. When the applied potential was 0.8 V, the surface also exhibited superoleophobicity and high oil adhesion underwater. Therefore, the underwater oil adhesion and wettability of the PANI membrane was controlled by adjusting the electrochemical potential. In terms of the internal mechanism of this phenomenon, the PANI nanowire film was in ES state when the external potential is 0.43 V, and protonation occurred at the nitrogen atom of the imine (generating polysemiquinone), thus the anion $({\mathrm{ClO}}_4{}^{-})$ had to act as a counter ion at the N⁺ position to maintain charge neutrality. Therefore, the interaction between the surface and water molecules was enhanced, and the microstructure of the surface could be completely wetted by the electrolyte solution, giving the surface a high OCA and a low oil adhesion. But when the external potential decreased to −0.2 V or increased to 0.8 V, the PANI form changed from the ES state to the LEB or PNB state, and ${\mathrm{ClO}}_4{}^{-}$ was removed from the main chain of PANI, so that the interaction between PANI and water molecules was weakened. Consequently, the oil partially penetrated into the microstructure of surface, resulting in a high oil adhesion.

Tian et al. [94] showed a method of directionally driving underwater oil droplets by jointly controlling the surface (gradient structured porous polystyrene film) through the electric field and the gradient microstructure (figure 8c). As shown in figure 8d, the micropores on the surface gradually decreased from one edge to the others, showing a gradient distribution, and the corresponding underwater value of OCA increased from 34° to 45°. They placed liquid paraffin on the gradient surface immersed in water, where an unbalanced pressure force caused by the gradient of the membrane was generated at both ends of the droplet. This unbalanced pressure further made the shape of the two sides of the oil droplet asymmetric, having a specific tendency to move. When an electric field was applied, the contact area between the underwater oil droplet and the gradient structure porous membrane was reduced, thereby overcoming the original viscosity resistance of the oil droplet and the prepared surface. When the pressure difference generated by the gradient was greater than the viscous resistance of the membrane to the oil droplets, the underwater oil droplets could move from the large pore area on the surface to the small pore area (figure 8e).

Liu et al. [101] developed a simple redox reaction using conductive polypyrrole (PPy) film to realize the intelligent switching between the high and low adhesion in terms of underwater oil droplets on the surface, and a potential wetting switch based on the oil/water/solid system was proposed. When a positive voltage was applied to the material, oil droplets adhered to the PPy surface. Conversely, when a negative voltage was applied, the adhesion force of the oil droplets on the surface of the material decreased, so it starts to roll under the influence of gravity, showing different droplet behaviour.

Finally, Yan et al. [83] revealed a unidirectional manipulation strategy for underwater oil droplets or bubbles on the surface of copper wire coated with porous polystyrene (PS) film by combining electrowetting induced adhesion conversion and the gravity/buoyancy of oil droplets/bubbles (figure 8f). The as-prepared porous PS film-coated copper wire was able to capture and maintain oil droplets or bubbles underwater without applying voltage. When the applied voltage increased, the wettability of the surface by water also increased, so the adhesion of underwater oil droplets or bubbles on the surface was decreased (figure 8g,h). Depending on the relative density, underwater oil droplets/bubbles could move unidirectionally downward/upward on the surface at a voltage greater than their critical voltage (figure 8i). Otherwise, underwater oil droplets or bubbles firmly adhered to the surface of the porous PS membrane (figure 8i).

3.5. Photo-responsive surfaces

3.5.1. Atmospheric environment

The studies on photo-responsive surfaces operating in either atmospheric or underwater environments have received increasing attention over recent years. Photo-responsive surfaces not only have some traditional applications (e.g. fog harvesting, microfluidics, liquid transport, water/oil separation), but also could potentially be used for boarder applications such as for photocatalysis and as microreactors.

The photo response (figure 9) can realize the remote control of surface wettability and adhesion without direct contact or introducing electrodes [107,108]. Therefore, compared with other responsive surfaces, more modes of switch in wettability and adhesion can be realized (table 6). Furthermore, both inorganic and organic materials (table 6) can be designed to be a photo-responsive surface, but each will have a different mechanism.

Figure 9.

Photo response or light driving on surfaces (in air). (a–c) On a ZnO ALD-coated surface, WCA switch between superhydrophobicity and nearly superhydrophilicity. Adapted from [102]. Copyright © 2010, American Chemical Society. (d) On a ZnO-composite SLIPS, CA switch without SA change. Adapted from [103]. Copyright © 2019, Wiley-VCH. (e–g) On a ZnO nanoparticle suspended surface, CA or/and adhesive force switch among superhydrophobic low adhesion, superhydrophobic high adhesion, and hydrophilic. Adapted from [104]. Copyright © 2014, Elsevier Ltd. (h–i) On an azobenzene-contained microsphere array structured surface, adhesion property switch from lotus leaf state to rice leaf state. Adapted from [105]. Copyright © 2018, American Chemical Society. (j,k) On an Fe₃O₄ nanoparticles embed lubricant-infused PDMS organogel, droplet manipulation due to photothermal response induced retention force difference. Adapted from [106]. Copyright © 2018, Wiley-VCH.

Table 6.

Advantages and diversity of photo-responsive/light-driven surfaces (in air and underwater).

advantages	details	ref.
diverse types of materials of interface can be used	nano metallic oxides/semiconductor, e.g. ZnO, TiO2, SnO2	[102–104,109–119]
	photoactive polymers e.g. azobenzene-based polymer	[105,106]
all modes of switch in surface properties can be realized	SA change as CA switch, switching between (super-)lyophilic and slippery state	[102,114,120,121]
	CA switch without SA change	[103]
	SA switch without CA change, switching between pinning and slippery state	[119]
	switching among lyophilic, slippery and pinning state	[104]
	switching between isotropic and anisotropic state	[105]
diverse control avenues are feasible	UV illumination	[102,114,115,117,120,122,123]
	UV light through a photomask	[104]
	LPL irradiation	[105]
	infrared-light irradiation	[106]
	varying the distribution, duration or the power of light irradiation	[105,106,118]
	to switch back: put in the dark	[114]
	to switch back: heat treatment	[104,116]
	to switch back: chemical treatment	[117]

Similar to the mechanism of semiconductor materials, the surface energy of various nano metallic oxides can be easily changed by illumination [108]. For example, ZnO nanoarrays were formed on the materials, where the surface micro/nano-structures were changed and the photo response under ultraviolet (UV) illumination was achieved [120,122,123]. Such a response could reversibly switch the wettability between superhydrophobic and superhydrophilic [120,121]. Researchers also found that TiO₂, SnO₂ could possess similar performance [109–112]. Contrary to conventional nanoarray decoration, Malm et al. [102] developed an atomic layer deposition method, by which an ultrathin ZnO layer was coated on a superhydrophobic cicada wings. This coating made minimal variation to the surface topography but remarkably changed the surface chemistry. The WCA on such a ZnO coating is photo-responsive. As shown in figure 9a, the surface maintained a superhydrophobicity after the ZnO coating. Under UV illumination, the WCA decreased sharply from around 160° to 14° (figure 9b,c). The amplitude of the CA variation was also controllable by optimizing the ZnO deposition cycle (figure 9c). Such a hydrophilic ZnO coating could recover superhydrophobicity when stored at 100°C in the absence of light. The above studies were based on the observation that a wettability switch brings about an associated adhesion switch. Recently, Han et al. [103,113] developed a ZnO composite SLIPS which had a photo-electric cooperative response. On such a surface, could they realize a reversible switch of CA without an SA change, as shown in figure 9d.

The surface adhesion property switch between superhydrophobic low and high adhesion could also be controlled on photo-responsive metallic materials [119]. Li et al. [104] sprayed a ZnO nanoparticle suspension onto a substrate. As shown in figure 9e, the WCA was 162° and SA was 2° after spraying, showing a superhydrophobic low adhesion property. When the surface was irradiated by UV light through a photomask, it transformed from low adhesion to very high adhesion where the droplet was pinned, without changing the superhydrophobic property (figure 9f). The corresponding adhesion force ranged from 5.1 to 136.1 µN (figure 9e,f). When the surface was under UV illumination without a mask, it became a hydrophilic surface (figure 9g). Through UV irradiation and heat treatment, the reversible transformation among superhydrophobic low adhesion, superhydrophobic high adhesion and hydrophilic could be realized. Velayi et al. [116] further studied how the annealing temperature in the heat treatment affected the switch in wettability and adhesion property.

Another approach to studying photo-responsive surfaces which has developed rapidly in recent years has been to employ photoactive polymers. The photo-response of polymers results from chromophore photoisomerization or the photothermal effect [107,124–130]. Azobenzene is a typical example of photochromic molecule [131,132]. The shape of a polymer containing azobenzene can be varied under different lighting conditions [133–135]. For example, Gao et al. [105] decorated the surface of substrates with photo-responsive poly[6-(4-methoxy-4′-oxyazobenzene)hexyl methacrylate] microsphere arrays. The substrate was initially treated to a lotus leaf state with isotropic superhydrophobicity. After irradiation using linearly polarized light (LPL), the microspheres transformed into an ellipsoid shape. As shown in figure 9h, the surface remained highly hydrophobic but exhibited anisotropic adhesion. The WSA for droplets moving parallel to the longer axes of the ellipsoid was 3° but the SA for moving vertically was 18°. As shown in figure 9h,i, enhancing the exposure time to LPL increased the degree of anisotropy of the adhesion so that the surface could change from lotus leaf state to rice leaf state. They explained the adhesion change by TPLC switching from the totally disrupted state to the anisotropic continuity state. Additionally, polymers which contain anthracene, stilbene, acetoxystyrene are also good candidates for photo response smart surface [107,136,137].

By adding photothermal substances into polymers they can respond under light due to the photothermal effect [138]. For instance, Gao et al. [106] embedded photothermal Fe₃O₄ nanoparticles in a lubricant-infused PDMS organogel and achieved a photothermal responsive smart surface. When a local infrared-light irradiation was exerted on the surface, a dynamic temperature gradient formed, as shown in figure 9j. The gradient gave rise to an imbalance between the advancing and receding angle so that the droplet could move (figure 9j). By varying the distribution and the power of light irradiation, the retention force against droplet moving towards different direction was optimized artificially, enabling the water droplets to move to a desired location at the desired speed or acceleration (figure 9k).

3.5.2. Underwater environment

As mentioned above, UV light could finally affect the interaction of some water/solid interfaces. When it comes to the underwater environment, the influence on water of surface structures or chemistry change is linked to the water/oil/solid interface, which exerts a significant effect on the oil droplet behaviour (figure 10).

Figure 10.

Photo response on smart surfaces (underwater). (a–c) TiO₂ surface. Adapted from [118]. Copyright © 2015, The Royal Society. (a) SEM images of the rough Ti surface after femtosecond laser ablation. (b) Underwater oil droplet on the rough sample surface after (i) storing in the dark, (ii) irradiation with UV light and (iii) Alternation of the above dark and UV light, respectively. (c) Schematic of the switchable wettability. (d) Schematic of large-scale fabrication process of the As-Spun nano textile and as well as the schematic of the photo/heat-induced wetting behaviour for this surface. Adapted from [115]. Copyright © 2020, American Chemical Society.

For example, Yong et al. [118] constructed a new type of multi-layered TiO₂ surface with switchable wettability by femtosecond laser etching of titanium plates (figure 10a). After being immersed in water, the surface reflected uniform light in the TiO₂ area like a silver mirror. This phenomenon was caused by the air band between the water and the substrate [139]. Oil droplets quickly spread over the underwater interface when they touched the rough structure (OCA = 4°), which indicated that the laser-treated TiO₂ surface was underwater superoleophilic (figure 10b). After UV irradiation for 40 min, the surface exhibited totally different wettability (OCA = 160.5° ± 2°). Meanwhile, the silver mirror-like reflected light disappeared, showing that water replaced the original air band to wet and fill the gaps of micro/nano-structures on the TiO₂ surface. The filled water layer was repellent to oil, which enabled the surface to exhibit superoleophobicity underwater. They also found that after placing the sample in dark air for 2 days, the initial underwater superoleophilicity reappeared; the above two states had repeatable characteristics (figure 10b). This reversible switch was based on the UV-sensitivity of TiO₂ [114,140]. When TiO₂ surface was exposed to UV light, a large number of electron pairs and holes were formed on the surface. The generated holes would further combine with the lattice oxygen to form highly unstable oxygen vacancies. Each oxygen vacancy dissociated water molecules to generate two adjacent Ti–OH groups. Owing to the hydrophilicity of the –OH group, water was able to completely wet microstructures when UV-treated TiO₂ surface was immersed in water. When the surface was placed in dark air, the hydroxyl groups generated by the dissociation of oxygen vacancies were replaced by oxygen in the air, restoring the surface to its original underwater superoleophilicity (figure 10c) [114].

Wang et al. [117] used the hydrothermal method on a fluorine-doped tin oxide substrate to fabricate a uniformly distributed rutile-type TiO₂ flower with a layered structure in a diameter of about 6 µm. The structured surface exhibited superoleophobicity in water (OCA = 155°). The surface had a low oil adhesion due to the formation of a fish scale-like solid/water/oil system in which water molecules were trapped in the TiO₂ microstructures and formed a water layer showing oil repellency. However, when the sample was contaminated with oleic acid (e.g. 0.5 wt% in n-pentane for 5 s), it lost the underwater superoleophobicity (the OCA decreased to 64°). This was mainly because during the process of oleic acid contamination, some oleic acid spots were also trapped in the microstructure of the TiO₂ surface, so that the droplets on the TiO₂ surface were in a non-uniform adhesion state. The oleic acid spots in the microstructure caused the reduction of contact area between the water and the oil droplets in the microstructure, thus decreasing the underwater OCA. Interestingly, when the surface was irradiated with UV light for 2 h, the original underwater superoleophobicity recovered. The recovery occurred because electron-hole pairs generated on the UV-treated TiO₂ surface. These holes adsorbed water molecules around the interface and produced hydroxyl radicals that were highly susceptible to reaction. The hydroxyl groups further oxidized and decomposed organic compounds such as oleic acid, which made TiO₂ surface return to superhydrophobic.

Shami et al. [115] further reported a three-dimensional reversible wettability PVDF-P25TiO₂ nano-textured surface, which was prepared by electrospinning using PVDF and P25TiO₂ as raw materials (figure 10d). The light-induced superhydrophilic and underwater oleophobic properties of the composite surface played an important role in the oil/water separation. During the separation process, the water in the mixed liquid was able to quickly pass through the surface due to the superhydrophilic property of the surface. When the oil reached the surface, it would be blocked at one side of the surface because of the underwater superoleophobic property, achieving water separation (figure 10d). In addition, the water on the surface could be removed by a simple heat treatment after the separation operation, therefore the surface was reusable.

3.6. Humidity-responsive surfaces

Humidity can be considered a dependent variable which can affect droplet behaviour. In nature, many plants respond to the changes in humidity by adjusting either their water absorption or evaporation. [141–143] On some specific bioinspired micro/nano-structured surfaces, the surface wettability and adhesion may also vary when they are exposed to different levels of humidity.

Recently, Qu et al. [144] discovered that the porous nano-structure of a peanut leaf could automatically respond to humidity changes (figure 11a–c). When the ambient relative humidity was lower than the saturated humidity (namely RH < 99%), as shown in figure 11a, the surface showed a superhydrophobic high adhesion property, with a WCA approximately 150° and an adhesive force of about 40 µN (figure 11c). When the RH > 9%, the surface suddenly switched to a superhydrophilic state (figure 11b) owing to the humid air filling the nanoscale framework. The WCA decreased dramatically to below 10°, showing a switch from superhydrophobicity to superhydrophilicity. Inspired by the peanut leaf, they fabricated a humidity responsive metal-organic framework (MOF) which had a similar performance (figure 11d–f). This smart surface also showed a switch from superoleophilicity to superoleophobicity/underwater superoleophobicity (figure 11f) due to the formation of a water film when the RH increased.

Figure 11.

Humidity response on smart materials (in air). (a–c) On a peanut leaf, WCA switch between superhydrophobicity and superhydrophilicity. (d–f) On a peanut leaf-inspired MOF, water and OCA switch between superhydrophobicity/superoleophilicity and superhydrophilicity/superoleophobicity. Adapted from [144]. Copyright © 2020, American Chemical Society. (g–i) On a nanocone-decorated three-dimensional multi-crossing fibre network, surface adhesion property switch from high retention force to low retention force. Adapted from [145]. Copyright © 2019, American Chemical Society.

In the same way that the humidity responsive formation of a water film can enable a surface to be more oleophobic and hydrophilic with high water adhesion, the break of a water film can also relate to the switch in the surface adhesion property. Recently, Li et al. [145] developed a ZnO-nanocone-decorated three-dimensional multi-intersectional fibre network (figure 11g) and found that the surface adhesion switched during the application of fog-water harvesting. As shown in figure 11h, the original web was hydrophobic with medium water retention force (around 800 µN). After hydrophilic ZnO-nanocone decoration, the web was more attractive to water and the retention force against water gathering become higher. When the web was exposed to humid air, the retention force decreased dramatically from 2090 µN to 230 µN after being wetted, accelerating the water movement for collection. They explained this smart switch by a humidity induced Rayleigh instability break in the water film (figure 11i): the unique web which had three-dimensional multi-intersectional structure together with hydrophilic nanocone array decoration [146–148].

The humidity response induced switch in surface wettability, adhesion and droplet behaviour is potentially an exciting research field, where on the interface the water (droplet) is affected by water (moisture). In addition, the underwater state could be regarded as an extremely humid condition, where the moisture/water exert maximum influence upon the interface.

3.7. pH-responsive surfaces

A variation in pH can also change the underwater surface wetting and oil adhesion. pH-responsive surfaces could adjust the contact conditions by the solution's pH, which allows the surface to effectively capture and release the oil droplet in solution. This makes it applicable to the field of marine oil pollution treatment.

Cheng et al. [149] prepared a surface that was able to adjust the wettability of underwater oil through pH by assembling HS(CH₂)₉CH₃ and HS(CH₂)₁₀COOH on the substrate surface (figure 12a). The surface showed superoleophilicity in a neutral or acidic aqueous solution (pH ≤ 7), on which the underwater OCA (octane, 4 µl) was constant at 0°. But when the surface is placed in an alkaline solution, its underwater OCA value increased as the pH. When the pH of the solution reached 12, the surface showed superoleophobicity (OCA = 162°) in the water (figure 12b). Experiments found that after the sample surface was removed from the alkaline solution and washed with deionized water, and then immersed in an acidic liquid again, it would return to underwater superoleophilicity. This reversible transition between the underwater superoleophilic state and the superoleophobic state can be repeated multiple times. The inherent reason for the above switchable wettability was the uniform distribution of carboxylic acid groups and alkyl groups on the surface of the sample, where the carboxylic acid groups would be protonated in acidic solutions and deprotonated in alkaline solutions. Therefore, when the surface was placed in a neutral or acidic solution (pH ≤ 7), the alkyl group on the sample surface dominated, enabling the surface to have lower surface free energy, so it showed superoleophilicity in water. By contrast, under alkaline solutions, the surface carboxylic acid groups are deprotonated (figure 12c). The deprotonated state was hydrophilic, allowing water molecules to be trapped in the rough structure. The higher the pH, the higher the degree of deprotonation of the carboxylic acid group. When the pH of the solution was higher than 12, the surface microstructure was able to capture a sufficiently thick water layer, thereby forming a stable solid/water/oil three-phase system leading to underwater superoleophobicity.

Figure 12.

pH response and the liquid behaviour (underwater) on surfaces. (a–c) HS(CH₂)₉CH₃ and HS(CH₂)₁₀COOH composite surface. Adapted from [149]. Copyright © 2014, American Chemical Society. (a) SEM images of the artificial surface assembled of HS(CH₂)₉CH₃ and HS(CH₂)₁₀COOH. (b) Wettability of oil droplet in acidic (pH = 2) and basic (pH = 12) solution, respectively. (c) Schematic of the responsive oil wettability on the rough surface. (d–g) PPFPA-g-PMAA surface. Adapted from [150]. Copyright © 2020, Wiley-VCH. (d) Schematic of fabrication procedures. (e) FE-SEM and magnified images of PPFPA micropillars and PPFPA-g-PMAA. (f) Underwater oil contact angle (UOCA) values and corresponding images of pillar-PPFPA-g-PMAA surface at different pH values. (g) Schematic of underwater oil droplet on the pillar-PPFPA-g-PMAA surface.

Inspired by the micro/nano-structure of fish scale, Huang et al. [150] prepared a pH-responsive underwater superoleophobic film by combining in situ anodic aluminium oxide (AAO) template polymerization and polymerization modification (figure 12d). The surface of the film was composed of polypentafluorophenyl acrylate micropillars grafted with polymethacrylic acid nano-brushes. The structure enabled the film surface to have a micro/nanoscale layered structure, which not only kept the surface superoleophobic in water, but also enabled the surface to adjust oil adhesion (figure 12e,f). Under acidic conditions, oil droplets could remain on the surface and exhibited the pinning state. If the solution turned to neutral or alkaline conditions, the oil droplets on the surface would turn into the slippery state where free sliding of oil drops occurred on the surface (figure 12g). Based on the underwater sliding oil angle in response to pH, the surface could be reversibly switched between low and high oil tightness.

Zhang et al. [151] designed underwater smart surfaces by grafting pH-responsive P2VP-b-PDMS (P2VP: Poly(2-Vinylpyridine)) on non-woven textiles and sponges. For the grafted P2VP block, it could change surface wettability and conformation through protonation and deprotonation in response to the pH value of the aqueous medium. The oleophilic PDMS block grafted on the surface could control the oil droplet entry, and further provided a controlled oil entry route through the oleophilic PDMS block, thereby regulating the oil wettability of the surface in water.

In addition, Cheng et al. [152] grafted a pH-responsive polyacrylic acid (PAA) layer on a glass substrate by a plasma polymerization method. The underwater adhesion of the grafted PAA film to oil droplets was regulated by the pH of the solution. As the pH gradually increased from 1.0 to 12.0, the adhesion of the surface to underwater oil droplets gradually decreased from 21.6 ± 5.0 µN to 0 µN. If the surface was placed in a low pH solution again at this time, the adhesion between the surface and the oil increased again, and the switching between the different states described above was repetitive. They also found that the adjustable oil adhesion possessed by the surface was caused by the chemical composition (pH-dependent) and micro/nano roughness of the surface. When the pH of the solution was less than 4.7, adjacent carboxylic acid groups on the PAA chain could form intramolecular hydrogen bonds, and the conformation of the PAA molecule was in a coiled state, thus the proportion of water on surface became low. The oil droplets could then be in contact with the surface rough microstructure. In this case, the surface had an underwater OCA of less than 150° and an extremely high adhesion to oil droplets. When the pH of solution was higher than 4.7, the intermolecular hydrogen bonds were formed between the carboxylic acid group on the PAA branch and the water molecules in the solution, and the conformation of the PAA molecule became stretched. Also, the intermolecular hydrogen bond made the proportion of water on surface higher, forming a water layer between the surface and the oil droplet, which provides a repulsive force for the oil showing high adhesion.

We can know that all the above pH-sensitive surfaces are grafted with one or more polymer materials that could respond to pH. When the pH of the solution changes, the functional groups/molecules on the surface also change, resulting in the variation of underwater wetting and adhesion of oil droplets.

4. Summary and outlook

Smart surfaces can respond to the stimuli of stretching, magnetic, photo, electric, temperature, humidity and pH. These smart surfaces use either a single response or multiple responses to these stimuli. Reviewing the studies on diverse stimuli-responsive switchable-adhesion surfaces working in atmospheric or underwater environment, the following conclusions may be drawn. Firstly, the contact conditions between a surface and a liquid can be divided into four states in either an atmospheric or underwater environment, i.e. lyophilic, pinning, slippery and anisotropic state. Responsive smart surfaces are able to reversibly switch from one contact state to at least one other state under the influence of external stimuli. Secondly, for artificial stimuli-responsive surfaces, the chemical component, chemical distribution, surface topography or/and parameter of environmental phase are changed when exposed to external stimuli. The wettability, adhesion and/or the contact state of the surface are therefore transformed, leading to changes in the liquid behaviour. By further varying the parameter of the external stimuli through a diverse control avenue, various modes of liquid behaviour can be realized, e.g. droplet deformation, unidirectional motion, bouncing and splitting.

Although great progress has been achieved in the development of responsive smart surfaces systems, many fundamental and practical challenges remain for this technology to find widespread application. As for potential future studies in this field, we propose different suggestions for some fields where smart surfaces are employed:

In terms of the research and application of accurate droplet manipulation, to fully understand the mechanisms of chemical or topographic effects on surface adhesion, researchers should ensure that the effects caused by other factors will not influence the results. For example, in the studies on temperature-responsive surfaces, micro/nano-structures on a surface were transformed when the surface was heated. However, the temperature of the liquid under investigation might also be changed, which might lead to a change in the properties of the liquid [153]. Researchers should minimize the influence of this change when evaluating how stimuli-induced nano-structure variation affects liquid behaviour.

Concerning the research on fluid mechanics and fluid control, the behaviour of other types of liquids (e.g. non-Newtonian fluids) [154] on the surfaces should be studied. In the majority of the current studies on liquid behaviour, only water and light oil have been studied systematically, both being Newtonian fluids. However, the non-Newtonian fluids (e.g. biofluid, some saline solutions, some polymeric solutions) show different behavioural trends in response to external forces, especially on hydrophilic surfaces [155]. Only when the behaviour of non-Newtonian fluids are widely investigated can more potential applications of smart surfaces be realized.

In terms of the application in the medical and pharmaceutical domains, designing soft wearable devices is a very promising avenue of exploitation. A challenge here is to link the ability of a surface to exhibit a liquid response with the ability of a surface to adhere to human skin. Most studies on smart wearable devices have been carried out when the skin is dry. Few studies have investigated how sweat or other liquid skin secretions affect the ability of surface to adhere to human skin [156]. A controllable responsive surface may be able to realize the ability when the skin is dry and also when the skin is not dry. In addition, we recommend that further studies which aim to promote smart surface applications in the biomedical fields should use biocompatible materials as substrates.

To promote the practical application of smart surfaces for liquid transport, fog harvesting, reactors or other purposes in hydraulic and chemical engineering fields, researchers should explore methods to prepare relatively large-scale surfaces at low cost. The durability of such a surface working in practical conditions should also be investigated.

To open the door to wider practical application, we recommend that stimuli-responsive surfaces can be combined with the concept of a bioinspired hybrid or the Janus interface [157], which possess wettability gradients. The artificial responsive surfaces reviewed here mostly had continuous wettability and switched between different continuous states. From our perspective, a wettability-gradient material can be employed as a substrate when a stimuli-responsive surface is designed. A stimuli-responsive hybrid/Janus interface may realize reversible switching in the direction or even the distribution of the gradient. Therefore, more factors related to the interfacial property and liquid behaviour can be more accurately controlled. Such a combination may lead to emerging applications such as smart clothing. For example, such an interface may be able to realize accurately controllable absorption and removal of sweat, heat and air, while it is hard for either single stimuli-responsive surfaces or single hybrid/Janus interfaces to achieve this.

Finally, we are convinced that designing artificial stimuli-responsive smart surfaces is a highly promising research and commercial field. The development of smart surfaces is a rapidly evolving with diversity. Various responses can be designed and diverse materials can be employed. Multifunctional surfaces can be made using common materials. As this review has highlighted, smart surfaces have potential applications in numerous fields, such as microfluidics manipulation, liquid transport/harvesting, thermal/energy management and flexible intelligent device design relating to the fields of biomedical devices, electronics, chemical engineering and many more.

Data accessibility

This article does not contain any additional data.

Authors' contributions

C.L. and M.L. conceived, designed and executed this project; they contribute equally to this work. Z.N. and Q.G. assist the writing and revision of the manuscript. B.B. and E.S. revised the language of the entire manuscript; they also provided useful suggestions for the revision, especially the summary and outlook section.

Competing interests

We declare we have no competing interests.

Funding

This work is funded by the EPSRC Program Manufacture Using Advanced Powder Processes (MAPP) EP/P006566. C.L. is supported by the China Scholarship Council (grant no. 202008060076). M.L. is supported by the President's Scholarship at Imperial College London. Q.G. is supported by the China Scholarship Council (grant no. 202006440011).