Introduction

Water-swelling is a widespread phenomenon characterized by the absorption of water molecules from the surrounding environment into the matrix of swellable materials, leading to volume expansion. This phenomenon is observed in nature, as seen in plant cells and soils exhibiting hygroscopic swelling when exposed to aqueous environments1,2. Additionally, it is a common occurrence in everyday technologies, such as kitchen sponges or disposable diapers3,4. The exploration of water-swelling phenomena has broad implications in fundamental science, and the development of functional materials. In this context, hydrogels have become exemplary artificial materials5. These are cross-linked polymer networks capable of absorbing up to approximately ten times their dry weight in water while maintaining structural integrity6,7,8,9. Hydrogels find applications in various fields, including drug delivery10,11, energy harvesting12, soft tissue scaffolds13, sensors14, soft actuators15, and soft robotics16. The remarkable “swelling” behaviors of hydrogels result from the hydrophilic nature of their polymeric matrix. The hydrophilicity arises from the presence of hydrophilic chemical residues17, such as hydroxylic (-OH)18, carboxylic (-COOH)19, and amidic groups, serving as binding sites that attract and retain water molecules, causing an increase in polymer volume20. It is important to note that the majority of reported instances of water “swelling” involve materials with inherent hydrophilicity. This raises a paradoxical question: Can water swelling be achieved using purely hydrophobic materials? Given that hydrophobic materials typically repel water and resist swelling, this question appears counterintuitive.

However, nature, with its diverse and ingenious designs, offers a compelling example in the form of Sphagnum, a plant that defies conventional expectations. In its desiccated state, Sphagnum exhibits hydrophobic characteristics due to its lignin-rich surface21. However, when immersed in water, this hydrophobic material undergoes rehydration, remarkably retaining water at a rate ten times its own weight22. With such excellent retention capabilities, Sphagnum is commonly utilized in horticulture to control water content in soil23,24. The intriguing phenomenon of the rehydration of hydrophobic Sphagnum is attributed to its inherent micron-scale porous structure, resembling a sponge with myriad tiny pores and spaces. The micron-scale porous structure can harness surface tension-induced capillary pressure to draw in water25,26. The surface tension, initiated by a small amount of water, cascades to attract more water, overcoming the inherent repelling forces of dried hydrophobic tissues (Fig. 1a and Supplementary Fig. 1).

Fig. 1: Schematic illustration of the hydrogel-like swelling behavior of the bioinspired sphagnum-structured porous elastomer for self-shaping and magnetically driven applications.
figure 1

a Depicts models of dried Sphagnum characterized by a porous surface facilitating efficient water absorption and storage. b Illustrates the schematic of a hydrophilic hydrogel, showcasing its absorption-induced expansion properties. c Presents a schematic representation of a hydrophobic elastomer immersed in water, maintaining its original state. d Demonstrates the absorption-induced expansion behavior of a porous hydrophobic elastomer after immersion in water. e Displays schematics of two composite systems: one combining porous elastomer with pure elastomers for programmable self-shaping transformations, and the other combining porous elastomers with magnetic elastomers for the preparation of magnetically driven robots.

Full size image

Inspired by the micron-scale porous structure of Sphagnum, we have engineered absorption-induced expansion materials using pure hydrophobic silicone elastomer endowed with a porous structure. This material, named hydrophobic pseudo-hydrogel (HPH), achieves unconventional absorption-induced expansion phenomena in a purely hydrophobic matrix through physically designed microstructures. We use the term “pseudo” here to indicate a resemblance to hydrogels in terms of water “swelling” ability (Fig. 1b), despite the fact that HPH is not a hydrogel as the underlying material is hydrophobic (Fig. 1c). In contrast to conventional hydrogels, where swelling involves molecular chain-level physicochemical interactions between water and the hydrophilic polymer backbone or lateral chains, the HPH utilizes capillary pressure and surface tension to accumulate water in microscopic pore structures (Fig. 1d). Our investigation observes this fascinating phenomenon and also provides a theoretical explanation. We propose a theoretical framework elucidating the interplay of elastocapillary and surface tension, establishing a foundation for understanding and controlling absorption-induced expansion behavior in hydrophobic matrices. We establish a correlation between expansion in this HPH and intricate pore structure, emphasizing pore size. Moreover, by systematically programming the pore structure within the elastomer, we demonstrate the realization of anisotropic and programmable absorption-induced expansion in this HPH. This programming capability extends to the achievement of dedicated and programmable self-shaping transformations27,28. Additionally, we combine magnetic particles with self-shaping HPH to prepare magnetically driven soft robots29,30 capable of basic behaviors such as swimming, rolling and walking (Fig. 1e). This counterintuitive material paves the way for fabricating adaptive, responsive soft materials, which can be used in soft actuation, implantation, flexible electronics, and other emerging materials applications.

Results and Discussions

To empirically validate our hypothesis, we developed a hydrophobic pseudo-hydrogel (HPH) using Ecoflex 00-30, a commercially available silicone rubber known for its hydrophobic properties. To induce a porous structure within the HPH, we employed a sacrificial template method using NaCl microparticles as the template, with details provided in Supplementary Fig. 2a at the “Methods” section. The cross-sectional morphology of the pristine silicone elastomer, as shown in Fig. 2a, was non-porous, and it exhibited a surface contact angle of 100° (Supplementary Fig. 3a), confirming its hydrophobic nature. Following the the particle leaching, the HPH displayed a porous structure, as illustrated in Fig. 2b, where the HPH was fabricated using NaCl particles approximately 60 μm in size. Here, the HPH thickness is around 1 mm, with a pore density controlled by incorporating 1:1 weight raito of NaCl microparticles into the silicone elastomer (Supplementary Fig. 2b). Additional samples with varying thicknesses (Supplementary Fig. 4), template contents (Supplementary Fig. 5) and pore structures (Supplementary Fig. 6) of HPH were also fabricated and investigated. It is crucial to note that the the particle leaching is a purely physical process, leaving the physicochemical properties of the elastomer unchanged. The surface contact angle of the HPH was 105° (Supplementary Fig. 3b), indicating that the physical porosity does not compromise the hydrophobic properties of the material itself. And the increase of the contact angle was attributed to the formation of porous microstructures on the elastomer surface31. Unlike the pristine elastomer, which did not swell when immersed in water for an extended period (six days) (Fig. 2c), the HPH, despite retaining its hydrophobic nature, unexpectedly undergoes absorption-induced expansion when immersed in water. As shown in Fig. 2d, after 132 h of immersion, the dimension of the HPH expands over twofold compared to its original size.

Fig. 2: Analysis of water absorption-induced expansion behavior in porous elastomers and the changes in mass and dimension during the expansion process.
figure 2

a Schematic of pure elastomer and SEM (Scanning Electron Microscope) image of a smooth cross-section. b Schematic of a porous elastomer and SEM image of a smooth cross-section. c The Optical picture of a pure elastomer sample in its original state after water immersion. d The Optical picture of absorption-induced expansion behavior of a porous elastomer sample after immersion in water. e Schematic during the wetting stage with large amount of air stored in the pore structure of HPH. f Schematic during the wetting stage with rarefied air entrapped in the pore structure of HPH. g Schematic during the expansion stage with rarefied air entrapped in HPH. h The curves of mass and volume change versus time of porous elastomer after immersed in water. i The expansion ratio versus time curves of porous elastomer with an average pore size of approximately 60 μm. Results are expressed as the mean ± SD for independent replicates (n  =  3).

Full size image

The absorption-induced expansion ability of HPH prompts us to reminiscent of conventional hydrogels. In the case of conventional hydrogels, their water “swelling” capability is rooted from their intrinsic hydrophilicity. These hydrogels typically consist of highly hydrophilic networks of polymer chains. Consequently, water serves as a thermodynamically compatible solvent for such hydrophilic networks, facilitating easy wetting of the dried hydrogel by permeating the hydrophilic polymeric network. Upon wetting, the hydrophilic polymer chains undergo solvation by water, creating an osmotic pressure gradient between the hydrogel and the surrounding aqueous environment. This gradient drives more water molecules from the surroundings into the hydrogel polymeric network, resulting in a macroscopic “swelling” phenomenon32,33. In contrast, HPH is a purely hydrophobic material, implying a distinct absorption-induced expansion mechanism compared to conventional hydrogels.

Inspired by the capillary-induced expansion observed in hydrophobic Sphagnum, we speculate that capillary pressure also play a pivotal role in the absorption-induced expansion of HPH. The absorption-induced expansion process of HPH can be delineated into wetting and expansion stages, similar to the “swelling” process of hydrogels, as schematically illustrated in Fig. 2e–g. Given the hydrophobic nature of HPH, water cannot readily penetrate its polymeric network for wetting. However, upon immersion of HPH in water, hydrostatic pressure or applying negative pressure (Supplementary Fig. 7) could compel water into the micron-scale porous structure of HPH. During this process, the air in the micropore structure was replaced by water, evidenced by the experimental observation of bubbles emerging from the HPH surface when immersed in water (Supplementary Fig. 8). Upon water introduction into the HPH, two competing capillary pressures arise: one along the liquid-elastomer curvature interface, which drives water further into the pores, and the other along the liquid-air interface, which opposes this process. As capillary pressure P is inversely proportional to pore radius R (following P∼2γ/R)34, the smaller radius at the liquid-elastomer interface generates a stronger net pressure, effectively driving water into the pores. This capillary pressure continues to progress, drawing water into the pore structures of HPH and concurrently expelling the originally entrapped air (Supplementary Fig. 9).

It is important to noted that, despite water entering the HPH during the wetting process, it merely replaces the volume of air within the HPH, without expanding the overall volume of the material. To elucidate this dynamic behavior of the wetting process, we plotted the volume change of HPH against time (t), as depicted in Fig. 2h. During the initial 30 minutes of being immersed, no significant changes in volume were observed. Concurrently, we monitored the mass change of the HPH, revealing a continuous increase. This weight increment indicates that the denser water is replacing the air inside the HPH. After approximately 30 min of immersion, the volume of the HPH begins to steadily increase, marking the commencement of the expansion stage. Due to capillary pressure, continuous imbibition is sustained, leading to the high stretchability of the elastomer and resulting in macroscopic volume expansion (Fig. 2g). Notably, although the deformation or expansion of the elastomer requires substantial pressure, the elastocapillarity at the liuquid/elastomer interface can generate sufficient pressure to induce deformation. Previous research on elastocapillarity has demonstrated that significant deformation in soft materials (elastomers) can occur when pore sizes are in the micron range, particularly below 100 μm35,36. With the pore structure of the HPH falling within this range, elastocapillarity effectively drives the observed expansion. As expansion progress, the elastomer’s expansion increases the radius of the pores, leading to a decrease in capillary pressure. On the other hand, the elastomer’s expansion generates an increasing elastic pressure that impedes water imbibition. Ultimately, a balance is achieved between the capillary pressure and the elastic pressure, resulting in the equilibrium expansion of HPH. As depicted in the volume curve (Fig. 2i), the absorption-induced expansion of HPH reaches an equilibrium state after 144 h of immersion. The absorption-induced expansion behavior of HPH is also evident in the microscopic pore structure changes. In Supplementary Fig. 10, a thin film of HPH ( ~ 78 μm) immersed in water is observed under a microscope, clearly revealing the pore structure. Approximately 90 min later, the pore sizes on the HPH surface progressively increase with the prolonged immersion time, eventually reaching equilibrium (Supplementary Fig. 11). The swelled HPH was also investigated by Environmental Scanning Electron Microscope (ESEM). As shown in the cross-sectional ESEM images of swelled HPH (Supplementary Fig. 12), the pore structure of HPH after absorption-induced expansion is demonstrated, showing that its pore size ranges from 80 to 200 μm. Moreover, comparing the pore morphology of the HPH before and after water-induced expansion, we observe that, prior to expansion, the pore shape is quite irregular, whereas after expansion, the pore shape is closer to spherical. This observation suggests an alternative explanation for the water-induced expansion in HPH through the lens of surface energy minimization. When water enters the irregularly shaped pores (Supplementary Fig. 12a), the liquid inclusions tend to adopt a spherical shape to minimize surface energy, as seen in previous studies36. This spherical tendency deforms the pore walls, thereby expanding them to approach a more spherical configuration (Supplementary Fig. 12b).

It should be noted here that, at the critical transition point between the wetting and expansion stages, a small portion of air remains sealed by the water inside the HPH (Fig. 2f), playing an important role in the subsequent expansion process. If all the air within the HPH were evacuated, leaving only the water/elastomer interface, the capillary pressure induced by surface tension around the pore periphery would be nullified, resulting in the absence of a driving pressure to draw water in. However, we consistently observe an increase in both volume and weight of the HPH. This sustained imbibition should be attributed to the trapped air, which generates asymmetric capillary pressure (Fig. 2g). The significance of the trapped air can be understood through the lens of surface energy minimization in water droplets. As discussed earlier, liquid inclusions strive to assume a spherical shape to minimize surface energy. The small amount of air trapped within the pores hinders perfect sphericity of the water inclusion. To approach a more spherical configuration, additional water is drawn into the pore, further expanding its structure. To confirm the presence of trapped air within the swelled HPH, the volume and weight curves were analyzed to calculate the density of the HPH. At the transition point, the density of the swelled HPH was 0.927 g/cm³, an intermediate value between the fully dried and fully wetted states. This suggests that at the transition point, about 11.69% of the pore volume is occupied by air (Supplementary Fig. 13), supporting the hypothesis of entrapped air. To futher demonstrate the importance of residual air in the pore for HPH expansion, we performed vertical absorption-induced expansion experiments on long strips of HPH. The HPH strip was positioned perpendicular to the surface of red ink, with their lower ends immersed in the ink (Supplementary Fig. 14). Unlike immersing HPH in water, where air is easily trapped in the hydrophobic matrix, this setup allows water to be transported upward from the bottom of the HPH, wetting the pores layer by layer. This process expelled most of the internal air from the HPH, leaving only a small portion to provide the driving pressure for the subsequent expansion process. As a result, the expansion ratio of HPH strips in the vertical absorption-induced expansion experiment is significantly lower than the one fully immersed in water. The observed expansion ratio in this setup is only 130%, demonstrating the crucial role of residual air in achieving full expansion.

As discussed earlier, the absorption-induced expansion process of HPH is linked to capillary pressure induced by liquid-elastomer surface tension along the periphery of the pore meniscus, which is also associated with the radius of the pores. Thus, by adjusting the pore size within HPH, the absorption-induced expansion ability of HPH becomes tunable. To explore the relationship between pore structure and elastomer expansion, the tunable pore sizes in the HPH were achieved by manipulating the size of the soluble NaCl particles. The internal pore structures of HPH with different pore sizes are illustrated in Supplementary Fig. 15. Observations reveal that smaller pore sizes correspond to a faster water absorption rate of HPH and a larger expansion size (Fig. 3a, b). Additionally, there exists a threshold pore size determining whether absorption-induced expansion will occur in HPH. Obvious expansion is observed only when the pore size is less than 152 μm. Conversely, when the pore size exceeds 152 μm, the pore structure in HPH is wetted but does not exhibit obvious expansion. This could be attributed to the fact that, when the pore size exceeds 152 μm, the capillary pressure generated by surface tension are insufficient to overcome the elastic pressure of the elastomer and induce expansion.

Fig. 3: The analysis of the mechanism of water absorption-induced expansion in HPH.
figure 3

a The expansion ratio versu time of HPH with different pore sizes. Results are expressed as the mean ± SD for independent replicates (n  =  3). b The optical image of HPH with different pore sizes before and after immersed in water. c Diagram of 2D ordered porous media with triangular arrangement of holes. d Porous media are filled with various phases of fluid. e Theoretical prediction results of between pore load modulus versus pore sizes. f Theoretical prediction results of pore equivalent Young’s modulus versus pore sizes.

Full size image

To further elucidate the intriguing pore size-related absorption-induced expansion phenomena of HPH, detailed principal calculations have been conducted to unveil the underlying mechanism, as shown in Fig. 3c, According to Gor et al.37, in porous media with regularly spaced pores, the engineering strain of a single thick-walled cylindrical unit cell can approximate the overall volumetric strain. To account for pore interaction, analysis is conducted on seven adjacent thick-walled cylindrical unit cells. In a single thick-walled cylindrical unit cell, pore walls endure internal pressure \({p}_{i}\), while the effect of neighboring pores on the central cell is akin to an external boundary pressure \({p}_{o}\). The absorption-induced expansion of HPH will provide the unbalanced pressure contributed by the surface tension near the liquid-solid interface, which will provide a driven pressure \({p}_{c}\) to break the balance between the internal pressure and external pressure. As Fig. 3d shows, before HPH wetting, \({p}_{i}={p}_{o}\), and the HPH will not swell. When we place the HPH in the water, this balance will be broken, and \({P}_{i}+{P}_{c}\gg {P}_{o}\). When pressurized liquid is injected into porous media with dual pore sizes, it undergoes two typical stages sequentially. Firstly, the macroscopic pores are rapidly filled, compressing the liquid inside the pores and exerting pressure on the pore walls, thereby stiffening the structure. Secondly, the pressurized liquid in the macroscopic pores gradually permeates into the microscopic pores within the framework under the driving pressure of pressure gradients, leading to complex changes in the overall macroscopic elastic behavior of the structure. Compared to the second stage, the first stage is typically completed in a very short time. Therefore, the focus here is mainly on the evolution of the macroscopic equivalent properties of porous media during the liquid infiltration process in the second stage. The HPH will swell very fast, thus, the pore pressure in porous media during the expansion will be\(\quad {P}_{f}=({P}_{i}+{P}_{c})-{P}_{o}\). In porous media containing fluids, there exist complex coupling effects between the solid framework and the pore fluid. The fluid pressure acting on the pore walls has a significant influence on their macroscopic mechanical behavior. A thorough understanding of the equivalent mechanical properties of fluid-containing porous media is crucial for promoting their practical engineering applications. Previous research has addressed this issue to some extent38. However, a systematic understanding is still lacking. Therefore, here, we analyze this problem based on a micromechanical model. Firstly, a theoretical model predicting the effective modulus of dry porous media will be established39, followed by an analysis of the influence of pore pressure. The pore load modulus in porous media can be expressed as

$${M}_{{pd}}=\frac{E\left\{\left[1+\left(1+\nu \right)k\right]-[1-(1+\nu )(1-2\nu )k]\xi \right\}}{\left(1+\nu \right)\left\{\left(1-2\nu \right)\left[1-\alpha \left(1+\left(1+v\right)k\right)\right]-[\alpha \left(1-\left(1+\nu \right)\left(1-2\nu \right)k\right)\xi -1]\right\}\xi }$$
(1)

Where \(k=(2\mu+\lambda )/({Ea})\), \(\mu\) and \(\lambda\) are the lame’s parameter, \(E\) is the modulus of the hydrogel, \(\nu\) is the Poisson’s ratio, \(\alpha\) is the porous ratio, \(\xi\) is the porous ratio-related parameter. The simulation details have been shown in Supplementary table 1, and the Young’s modulus of HPH with different pore sizes before and after expansion were shown in Supplementary table 2. In Fig. 3e, it can be observed that for nanoporous specimens with the same porosity, as the pore size decreases, the influence of surface effects becomes increasingly apparent. Thus, the equivalent volume modulus of porous media can be defined by the formula,

$${K}_{{pd}}=\frac{E(1-\xi )}{2[\left(1-\nu \right)+(1+\nu )\xi ](1-\alpha \xi )}$$
(2)

Where the \({K}_{{pd}}=\frac{{E}_{{pd}}}{2(1-{\nu }_{{pd}})}\), with \({\nu }_{{pd}}=1-[\left(1-\nu \right)+(1+\nu )\xi ]{\left(1-\xi \right)}^{2}\exp \left\{\dfrac{3\alpha \xi \left[2+\left(1+\alpha \right)\xi \right]}{2\left[2-\left(2-\alpha \right)\xi \right]}\right\}\).

And the equivalent Young’s modulus of porous media can be defined by the formula,

$${E}_{{pd}}=\frac{E{\left(1-\xi \right)}^{3}}{\left(1-\alpha \xi \right)}\exp \left\{\frac{3\alpha \xi \left[2+\left(1+\alpha \right)\xi \right]}{2\left[2-\left(2-\alpha \right)\xi \right]}\right\}$$
(3)

As shown in Fig. 3f, the equivalent Young’s modulus increases rapidly with increasing pore size and water swelling. This behavior is influenced not only by the reduced solid matrix but also by the interaction of water swelling within the porous media. When water infiltrates the pores, it induces swelling, which generates initial stress within the material to balance the surface tension forces. This process enhances the structure’s equivalent Young’s modulus by effectively stabilizing and reinforcing the material under these conditions. Consequently, as the equivalent Young’s modulus increases, the expansion of HPH becomes increasingly constrained. The theoretical model predicts a threshold pore size of approximately 180 μm, beyond which water absorption-driven expansion is significantly hindered. This aligns with our experimental observation that HPH shows no significant expansion when the pore size exceeds 215 μm.

Since HPH exhibits similar absorption-induced expansion ability to conventional hydrogels, it functions similarly and has similar potential applications. Conventional hydrogels have been extensively studied as self-morphing materials due to their water-triggered shape transformation phenomena. The development of self-morphing materials is inspired by the widespread shape-transformation phenomena observed in plants40, generating growing interest in various fields such as energy harvesting, metamaterials, soft robotics, sensors, and multifunctional bioscaffolds. In the following section, we explore the potential applications of HPH as programmable water-responsive self-morphing materials. Typically, the shape transformation of hydrogels arises from inhomogeneous absorption-induced expansion behaviors within the material. Traditional shape-morphing hydrogels achieve this by incorporating diverse components with different “swelling” behaviors in response to specific stimuli. However, introducing diverse components may encounter intrinsic limitations, such as unstable interfaces and time-consuming fabrication. Therefore, it is of great significance to develop monocomponent hydrogel that enable precisely programmable deformations41,42,43,44. Our HPH materials can take the advantages of the microstructure programmed absorption-induced expansion ability to realize monocomponent self-morphing materials (Supplementary Fig. 16).

By controlling the pore structural domain in HPH, programmable and localized expansion can be achieved, which then driving the controllable shape-morphing of HPH. The fabrication method is detailed in Supplementary Fig. 17. The porous domain in HPH (the domain with absorption-induced expansion properties) serve as the active component, while the non-porous domain, unresponsive to water stimulus, serve as passive component. The perfect bonding between the active and passive components is attributed to the fact that they are both Ecofelx 00-30 (Supplementary Fig. 18). Consequently, non-uniform volume changes generate internal stresses at their interfaces, inducing out-of-plane shape transformation. Additionally, the absorbed water in the swollen HPH matrix can be completely removed, allowing the HPH to return to its original state (Supplementary Fig. 19a). The swollen HPH can also maintain its shape and volume in air for up to 8 hours (Supplementary Fig. 19b, c), establishing a solid foundation for its application in portable deformable robots. Initially, we investigate the self-buckling deformation of HPH by laterally arranging porous (active) and nonporous (passive) components. As illustrated in Fig. 4a, a series of concentric patterns is programmed in HPH, where in-plane heterogeneous expansion leads to modulated internal stresses, resulting in 3D deformations. Upon immersion in water, the planar concentric circle HPH evolves into Enneper’s surfaces with controllable wrinkles. In Fig. 4a, we demonstrate patterned surfaces with three to six wrinkles. Finite element analysis (FEA) was utilized to predict the formation of these 3D structures. The simulation results closely matched the experimental data, highlighting the effectiveness of computational modeling in designing shape-morphing structures. For more intricate in-plane patterning, a periodically patterned HPH is prepared, featuring an array of non-swellable discs embedded in the swellable HPH framework. In this configuration, each compartmentalized swellable and porous domain is surrounded by four non-porous and non-swellable discs, leading to an alternating concave-convex 3D shape (Supplementary Fig. 20a).

Fig. 4: The application of HPH for humidity-responsive deformation.
figure 4

a The planar schematic, optical and post-deformation photographs of Enneper’s surfaces with controllable wrinkles. b The planar schematic, optical and post-deformation photographs of the folding structure. c The schematic of initial unit of the assembled lattice and the planar schematic, optical and post-deformation photographs of lattice structures.

Full size image

In addition to self-buckling shape transformation, HPH can also achieve self-bending and twisting (Fig. 4c) by arranging the porous structural gradient across the thickness of materials. Figure 4c provides several examples of expansion-induced 3D HPH achieved through unidirectional or bidirectional folding. More examples of expansion-induced shape transformation in HPH, which borrow techniques and ideas from conventional self-morphing hydrogels such as kirigami and responsive mechanical buckling, are demonstrated in Supplementary Fig. 20. Furthermore, a reconfigurable and assembled responsive lattice can be achieved. Additionally, as shown in Fig. 4d, we prepared a series of self-folding HPH strips and then assembled these strips into lattice by using 3D-printed dock connectors. These HPH strips can be assembled into various types of responsive lattice with different connection configurations as shown in Fig. 4d and Supplementary Fig. 21. These responsive lattices serve as a common platform for designing soft mechanical metamaterials capable of negative expansion ratios. Upon absorption-induced expansion, the lattice shown in Fig. 4d achieves a negative expansion with area change of 29% and 25%. In contrast to previous literature, our responsive lattice composed by HPH strips is dismountable with reassembling ability, enabling the recycling of self-folding HPH units into reconfigurable metamaterials. To further expand the potential applications, the fabrication of HPH structures using a dual-nozzle 3D printer was also explored. As demonstrated in the Supplementary Fig. 22, the materials are compatible with 3D printing, allowing for the creation of customized geometries. It was verified that the printed samples undergo shape deformation in water, underscoring the potential of these materials for 4D printing applications45,46,47.

Finally, like conventional hydrogels, HPH can be combined with functional components to achieve emergent actuation properties and performance30. Here, by incorporating magnetic NdFeB microparticles into the monocomponent self-morphing HPH, we report hybrid materials that are actuated by a magnetic field after expansion-induced shape transformation. For conventional hydrogels, before incorporating NdFeB, the NdFeB should undergo surface passivation, such as coating with a thin layer of silica, to prevent its corrosion in the matrix of hydrogel48. In contrast, our HPH is composed of the hydrophobic silicone elastomer, which is an ideal matrix for NdFeB and has been reported in many literatures. Incorporating 100 wt% NdFeB MPs into the silicone elastomer produces a magnetic HPH that can be actuated under an external magnetic field of 15 mT (Supplementary Fig. 23). As shown in Fig. 5a and Supplementary Fig. 24a, by harnessing the self-buckling deformation of HPH, we obtained a floating robot featuring a buckled buoy and a magnetic propelling tail. The buckled cavity provides enough buoyancy force to keep the robot floating on the water surface. Under magnetic actuation by an N52 magnet, the magnetic propelling tail flaps the water, propelling the floating robot forward (Fig. 5b and Supplementary Movie 1). The expansion-induced deformation not only changes the shape of the magnetic robots but also alters the magnetization profiles inside the materials. Before shape morphing, the magnetic HPH film is in a flat configuration, resulting in a 2D magnetization profile. After morphing into a 3D shape, the overall spatial arrangement of the embedded NdFeB particles changes due to the deformation of the film, although the relative positions of the magnetic particles within the silicone elastomer remain fixed. This leads to a change in the overall magnetization profile of the entire structure, reflecting the new 3D morphology of the HPH. Illustrated in Fig. 5c and Supplementary Fig. 24b, we initially design a flat magnetic HPH film with a simple planar magnetization profile. With this simple planar magnetization profile, it is challenging to realize magnetic actuation movement for the HPH film. After expansion-induced shape transformation, the HPH film rolls into a wheel-like structure. Moreover, such self-rolling shape transformation also deforms the original simple planar magnetization profiles in the HPH film, resulting in 3D axially divergent magnetization. With this more complex 3D magnetization and the wheel-like shape, the HPH film can achieve wheel rolling movement under magnetic actuation (Fig. 5d and Supplementary Movie 2). Figure 5e and Supplementary Fig. 24c exhibits another example by adopting self-morphing to tune the magnetization profile to realize delicate magnetic actuated movement. Our design involves a planar cross-shaped HPH film endowed with two magnetic “legs”. In the planar structure, these two “legs” cannot be used for walking. Upon expansion-induced shape transformation, these two magnetic “legs” can stand up to support the HPH body. Under magnetic actuation, the standing HPH robot can alternate its magnetic “legs” to walk forward (Fig. 5f and Supplementary Movie 3).

Fig. 5: The application of HPH in magnetic-driven robots.
figure 5

A schematic of the swimming robot (a) and time lapse optical photographs of swimming robot (2 s time interval) (b) in motion driven by a magnetic field. A schematic of the rolling robot (c) and time lapse optical photographs of rolling robot (4 s time interval) (d) in motion driven by a magnetic field. A schematic of the walking robot (e) and time lapse optical photographs of walking robot (nonconstant interval between images) (f) in motion driven by a magnetic field.

Full size image

In addition to incorporating magnetic functional materials, the integration of conductive functional materials was also explored to expand the capabilities of the hydrophobic pseudo-hydrogel (HPH). Specifically, conductive and flexible carbon fabric was embedded within the HPH to impart conductivity alongside its self-morphing capabilities. This conductive HPH demonstrates promising applications in 3D electronics and wearable electronics49,50. As shown in the Supplementary Fig. 25a, b, the shape-morphed HPH can serve as a 3D circuit to power an LED. Additionally, an adaptive sensor based on the conductive HPH was developed, which, after shape transformation, forms a ring-shaped structure (Supplementary Fig. 25c, e). This ring-shaped sensor was used to record electrocardiogram (ECG) signals by measuring resistance changes between the skin and the carbon fabric electrode, effectively monitoring the electrical activities of the heart.

In summary, our study presents an hydrophobic pseudo-hydrogel (HPH) inspired by the absorption-induced expansion phenomenon of hydrophobic Sphagnum. The HPH, despite its hydrophobic nature, exhibits unconventional absorption-induced expansion through surface tension-induced capillary pressure in its designed microstructures. The entrapped air during the wetting process plays a crucial role in sustaining continuous imbibition, leading to macroscopic volume expansion. Tunable absorption-induced expansion is achieved by manipulating pore size, highlighting its potential for diverse applications. The programmable self-morphing ability of HPH enables controlled shape transformations, offering versatility for soft robotics and reconfigurable metamaterials. Incorporating magnetic particles enhances actuation capabilities, exemplified by the creation of floating robots with distinctive movements. This hydrophobic material provides a unconventional avenue for designing adaptive, responsive soft materials with implications in various fields. The theoretical framework established in this study contributes to the understanding and control of absorption-induced expansion behavior in hydrophobic matrices, paving the way for future advancements in material science and engineering.

Methods

Materials: Part A and Part B of EcoflexTM 00-30 were purchased from Smooth-On, Inc. Sodium chloride (NaCl, 99.7%) was obtained from Sinopharm Chemistry Reagent Co. Ltd. (Shanghai, China). Magnetic nanoparticles (MQFP, NdFeB, 99.9%) were purchased from Magnequence. The N52 magnet was obtained from Jiahao Magnetic Products (Dongguan, 0.5 T). The purchased materials were used without further purification.

Characterization: SEM (Japan, Hitachi, SU8010) was used to observe the microstructure of the porous structure of the elastomer. Metallographic optical microscope (XK-200, Shenzhen Sinico Optical Instrument Co.) was used to observe the changes in the porous structure of the elastomer films with the time of absorption-induced expansion. The contact angle was measured by Contact Angle Measurement Instrument (JC2000D1, Powereach). The pore structure after HPH expansion was observed using an environmental scanning electron microscope (ESEM) (QYANTA200, FEI, American). Optical images of deformed HPH and the magnetic robots were taken using an smart phone. The actual magnetic field at the sample’s location was measured by a commercial magnetic sensor (Infineon TLE493D-W2B6).

Preparation of porous elastomer: NaCl was milled, sorted, and stored. Components A and B are mixed 1:1, with NaCl added based on the polymer-to-NaCl ratio. The mixture is cast into a PMMA mold, cross-linked at 60 °C for 30 min, then demolded, and washed with water to remove NaCl, and dried to form a porous structure.

Preparation of structurally complex 2D precursors: PMMA sheets are shaped with a CO2 laser to create structures. A stepwise template method forms 2D precursors as bases. An AB mixture with NaCl is infused into specific areas to create features, while other areas are filled with Ecoflex as a binder, forming 3D structures with buckling and folding patterns.

Preparation of magnetically driven soft robots: A similar template method to that used for porous elastomers was applied to fabricate magnetically driven robots. The process begins with mold design, followed by selecting the appropriate filling material for each section. The main part is filled with pure ecoflex, while other areas are shielded. After the ecoflex cured, the second part of screen is removed, and a NaCl-infused Ecoflex mixture is added. After curing, a third section is filled with a magnetic particle-Ecoflex mixture (1:1 weight ratio). The magnetic particles were pre-magnetized using an electromagnet (35 V, 4.8 A, with a pole gap of 1 cm) as shown in Supplementary Fig. 26. Layers are built up in this manner, and once fully cured, the elastomer is removed and immersed in water to dissolve the NaCl template, forming a porous magnetic driven elastomer robot.

Finite Element Simulation: A simplified finite element model of the elastomer was constructed using Abaqus 2020 (Dassault Systemes Simulia Corp., USA), and the deformation behavior of the elastomer was simulated using the Abaqus/Explicit solver. All models were defined as S4R shell elements with a mesh size of 0.5 mm, and the geometric parameters of the models were consistent with the measured dimensions. The mechanical and mass properties used in the simulations were derived from experimental measurements. Specifically, the elastic modulus of the pure elastomer was 0.065 MPa, while the porous had an elastic modulus of 0.045 MPa. The interface between the pure elastomer and porous elastomer was modeled as fully bonded. In the simulations, it was assumed that the elastic modulus remained constant during swelling. The bending deformation of the hydrophilic hydrogel was induced by applying a temperature gradient in the thickness direction of this region, with a gradient of 0.4 °C mm−1 and a linear thermal expansion coefficient set to 1 K–¹. Appropriate material damping settings were employed during the analysis to promote convergence.