Introduction

Clean water is the cornerstone of human health and industry1,2,3. However, water resources are not evenly distributed across the planet, and most exist as seawater that cannot be directly used1,4. Therefore, it is vital to use green energy to supplement water resources through seawater desalination and sewage purification5,6,7,8,9,10. As the representative of green energy, solar energy has been widely used in environmentally friendly and sustainable evaporation technology4,8,9,10,11,12. Solar energy is converted into heat energy to accelerate water evaporation at the water-air interface by the solar steam generators floating on the water5,13. The solar-thermal energy conversion mechanism mainly includes the local plasma effect of noble metal nanoparticles14,15, non-radiative relaxation of semiconductors4, thermal vibration of molecules in carbon-based materials16,17, etc. To enhance the solar-to-steam energy conversion efficiency, the solar evaporators are supposed to possess high photothermal conversion efficiency3,18,19,20, water management supplying consecutive water for continuous evaporation, and thermal management suppressing heat loss21,22,23,24. Compared with noble metal nanoparticles, carbon nanomaterials, such as reduced graphene oxide (rGO)18,21,25,26, carbon nanotube (CNT)17,26,27,28, MXene6,11,14,29,30, etc., possess excellent solar–thermal energy conversion properties in the whole solar wavelength. Meanwhile, the structure and composition of the carbon-based composites can be adjusted for integrated water-thermal management31,32,33. Therefore, carbon nanomaterials are widely used as precursors for efficient solar evaporators34,35,36.

By selecting raw materials, good photothermal conversion efficiency can be achieved, but it is difficult to coordinate water and heat management. In fact, the water involved in the evaporation of the solar steam system is only a tiny part of the volume near the water–air interface4,5,37,38. Thus, a consecutive flow of water is required to be transported to the air–water interface for continuous evaporation. However, thermal diffusion is inevitable during water transportation. There is always a part of solar-converted heat energy heating the water not directly involved in evaporation16,39,40. Therefore, coordinating water transportation to the air–water interface while minimizing heat diffusion is challenging11,17,41. Water management and thermal management conflict, and simple continuous micron vessels are challenging to meet both requirements. Therefore, to achieve integrated water-thermal management, it is necessary to design the evaporators’ microstructure meticulously. In previous research, bionics (feathers11, honeycomb19, mussel7, crumpled textures6,42, continuous vertical vessels29,41, etc.) is widely used to design the microstructure of the evaporators to realize the integrated management of water and heat. However, some issues still need to be addressed: 1. The solar steam generators cannot achieve passive integrated water-thermal management, which is not environmentally friendly13,39. 2. The evaporation performance of the evaporators under weak light conditions still needs to be improved19,41. 3. The structural design is not meticulous enough as part of the solar active nanomaterials (such as MXene, rGO, CNT, etc.) does not directly participate in solar energy absorption. Thus, the theoretical efficiency of the nanomaterials has been sacrificed25,29,31,42. 4. The bottom-up assembly makes the MXene nanosheets unnecessarily exposed to the water–oxygen environment during the process, resulting in the oxidization of MXene and thus limiting the working life of MXene-based evaporators43.

To tackle the above challenges, we fabricated the hierarchical MXene-rGO sponges (rGMXn sponges). The axial thermal conductivity of the sponges is designed to be lower than the radial one. Therefore, the heat is more likely to spread in the radial plane rather than in the axial direction far away from the air–water interface. The in-plane temperature of the sponges is more likely to be homogenized, effectively preventing the local steam pressure from reaching saturation6. Furthermore, the axial-oriented vessels in the sponges allow for continuous water transition. Thereby, the none-external-energy-input integrated water-thermal management is achieved.

In detail, the MXenes nanosheets are in situ synthesized perpendicularly on the surface of the sponges, which can maximize the exposure of MXene to the sunlight and acquire a high photothermal conversion efficiency. Besides, the rGMXn-based evaporator is obtained by in situ etching of the MAX-based sponge. Therefore, the oxidation of MXene during the bottom-up assembly is avoided, greatly extending the working life of the rGMXn-based evaporators. The low density, high hydrophilicity, and mechanical stability make the sponges float on the water as free-standing solar evaporators. Furthermore, the low density of sponges brings a very low thermal conductivity, effectively inhibiting the unnecessary diffusion of heat energy in water. The sponges obtained in this work have achieved water–thermal integrated management by anisotropic designing and possess good evaporation performance. The rGMX5 sponge with MXene content of only ~4 mg ml−1 can achieve an evaporation rate of 2.35 kg m−2 h−1 and an efficiency of up to 127% under one sunlight illustration, and the rGMX5 sponge maintains an evaporation rate of 1.6 kg m−2 h−1 with 85% energy efficiency even under weak sunlight (0.5-sun). Furthermore, the rGMX5 sponge possesses an extraordinarily long working life with a 96% retention of evaporation rate after sustained operation for 30 days.

Results and discussion

Fabrication of the MXene-reduced graphene oxide sponges

The MXene-sponge-based solar-driven evaporators are fabricated by in situ synthesis of MXene nanosheets on the surface of the rGO sponges (Fig. 1a, Supplementary Fig. 1)44,45. At first, precursor sponges are obtained by directional freezing and freeze-drying. Subsequently, by sacrificing part of rGO as the carbon source, a uniform MAX phase layer is synthesized in situ on the surface of the rGO sponges. Finally, rGMXn sponges are obtained with the MXene nanosheets vertically anchored on the rGO framework after in situ etching by acid. As there is no need for assembly after etching, the rGMXn sponge can be immediately put into use, greatly shortening the exposure time of MXene in the water–oxygen environment during the fabrication process. Besides, without the need for solvent displacement31, the MXene nanosheets are barely self-restacked, guaranteeing the utilization of MXene in the photothermal conversion of solar energy.

Fig. 1: Fabrication of rGMXn sponges.
figure 1

a Schematic illustration of the synthesis of rGMXn sponges. b Axial and c radial SEM images of rGMXn sponges. d XRD patterns of rGMXn and rGMAX5 sponges (n = 5, 10, and 20). e XPS Ti 2p spectra of rGMX5 sponge.

Full size image

In the rGMXn sponges, the rGO framework supports the structure of the sponges. The axial-oriented vessels provide fast water transmission channels for continuous evaporation at the air–water interface (Fig. 1b). Unidirectional freezing causes the sponge’s walls to have micro orientation and macro isotropy in the radial plane (Fig. 1c). Furthermore, the distribution of MXene nanosheets makes the radial thermal conductivity of the sponges much greater than that of the axial one. Therefore, the rGMXn sponges have achieved efficient integrated water and thermal management, ensuring to be potential solar-driven evaporators.

Inspired by the coral, the MXene nanosheets are vertically synthesized in situ on the surface of the rGO framework (Fig. 1b). With the growth of MXene nanosheets, the (002) peak of the graphene phase in the rGMXn sponges is gradually invisible in the X-ray diffraction (XRD) spectra, as the MXene nanosheets entirely covered the surface of the sponges (Fig. 1d). Besides, the disappearance of the 39° peak indicates that the aluminum in the MAX phase has been etched completely18,44,45. With the increased content of MXene, the ratio of D peak (A1g mode at ~1350 cm−1) to G peak (E2g mode at ~1590 cm−1) in the Raman spectra decreases, indicating the symmetry break of the microstructure of the sponges (Supplementary Fig. 2a)45,46. Besides, the peaks at 610, 410, and 250 cm−1 represent the nonstoichiometric δ-TiCx phases, indicating the MXenes in the rGMXn sponges are a combination of Ti2C and Ti3C247. The X-ray Photoelectron Spectroscopy (XPS) spectra have further proved this, as the Ti 2p XPS spectra (Fig. 1e, Supplementary Fig. 2) can be divided into four doublets located at 457.4 eV (reduced Ti ions), 454.5 eV (Ti\(-\)C bond), 455.9 eV (titanium oxy carbides and substoichiometric TiCx, x < 1), and 458.8 eV (Ti4+ ions)26,48.

Anisotropic mechanical performance of the sponges

The rGMXn sponges obtained by directional freezing possess low density (Supplementary Table 1) but strong mechanical stability and anisotropic mechanical performance. The rGMX5 sponge can support a thousandfold-dead-weight weight in the axial direction without macroscopic deformation (parallel to the freezing direction). At the same time, it is easy to be compressed in the radial direction, showing good elasticity (perpendicular to the freezing direction, Fig. 2a). The rGMXn sponges’ modulus along the axial direction is much greater than that in the radial direction due to the ice crystal’s growth during oriented freezing (Fig. 2b, Supplementary Fig. 3a, b). Furthermore, the rGMXn sponge has a more extensive elastic deformation range in the radial direction with stable cyclic performance (Fig. 2c, Supplementary Fig. 3c–f).

Fig. 2: Mechanical stability of rGMXn sponges.
figure 2

a Digital image of anisotropic rGMX5 sponge pressed by 200 g weight from different directions. b Anisotropic compressive modulus of rGMXn sponges. c Mechanical stability performance of rGMX10 sponges in the radial direction. d Digital images of rGMX5 sponges during water immersion. e Schematic of the anisotropic mechanical mechanism of the sponges.

Full size image

The significant anisotropic mechanical performance and strong mechanical stability can be expressed as follows (Fig. 2e): The oriented rGO framework provides excellent compressive strength in the axial direction, while the axial-oriented vessels ensure the sponges with ample space for compression in the radial direction. Furthermore, the van der Waals force between the MXene sheets vertically anchored on the rGO framework offers the sponges good elasticity16,29,49. It is worth pointing out that the modulus of the rGMXn sponges drops with the increase of MXene content (Fig. 2b, Supplementary Fig. 3a, b), which is due to the synthesis mechanism: The in situ synthesis of the MAX phase has sacrificed part of the rGO as carbon source, which has introduced a lot of defects in the framework45,46,48. The sponges can self-float on the water because of their excellent mechanical stability, low density, and good hydrophilicity. The stable mechanical properties of the rGMXn sponges can ensure long-term operation and maintain their integrity after immersion in water. Due to the anisotropic modulus, the rGMXn sponges can shrink in the radial direction when losing water and expand back to their original volume, quickly responding to water absorption (Fig. 2d). The structure of the sponge is very stable, which can be well maintained after rapid water absorption cycles (Supplementary Video 1).

Anisotropic thermal performance of the sponges

The rGMXn sponge has excellent thermal insulation performance due to its abundant porous structure18,50. Separated by a one-cm-thick sponge over the baking of the alcohol lamp, the plantlets can remain upright within 60 s (Fig. 3a, Supplementary Fig. 6a, b). As the MXene content increases, the thermal conductivity of the sponge increases. The rGMX5 sponge has an axial thermal conductivity of 0.025 W m−1 K−1 (Fig. 3b), while its’ radial thermal conductivity (0.268 W m−1 K−1) is tenfold higher. The thermal conductivity demonstration experiment (Fig. 3c) intuitively reveals that the increase of MXene content has little effect on the axial thermal conductivity (Supplementary Fig. 4a) but dramatically improves the radial thermal conductivity (Supplementary Fig. 4b).

Fig. 3: Anisotropic thermal performance of rGMXn sponges.
figure 3

a Digital image of a plantlet over an alcohol lamp flame separated by a thin rGMX5 sponge. Anisotropic b thermal conductivity, c infrared thermal images, and d a schematic of the thermal mechanism of the sponges.

Full size image

The smaller thermal conductivity of the rGMXn sponges with anisotropy compared to rGO sponges can be explained as (Fig. 3d): 1. Due to directional freezing, the holes in the sponge are directionally arranged, resulting in anisotropic thermal conductivity of the sponge44,51. 2. The heat transfer of sponges along the axial direction mainly comes from the heat conduction in the oriented rGO framework, as the continuous rGO framework possesses good theoretical thermal conductivity. However, the in-situ MAX phase growth has sacrificed a portion of rGO as the carbon source, disrupting the continuity of the framework. The heat conduction of the sponges along the framework is hindered. Therefore, the axial thermal conductivity of the rGMXn sponge slumps and is lower than other aerogels (Supplementary Fig. 5a)11,16,51. 3. The MXene sheets are perpendicular to the rGO framework, making it easy for heat to transfer from in-plane rGO to out of plane MXene with larger theoretical thermal conductivity, further decreasing the axial thermal conductivity of the rGMXn sponges. Meanwhile, the hierarchical structure has reduced the interfacial thermal resistance along the radial direction6,14. Therefore, as the MXene content increases, the radial thermal conductivity of the sponge increases.

The impacts of structural design on sponge evaporation performance

The outstanding absorption efficiency of sunlight is the prerequisite for achieving high solar-driven steam generation4,5,6. After etching, the light absorption of rGMXn sponges is greatly enhanced in the whole solar band, as the etching has increased the layer spacing of MXene sheets and the surface areas exposed to light (Supplementary Fig. 7a). Besides, the light absorption of sponges enhances with the increase of MXene content (Fig. 4a). When the light is incident from the axial direction, most of the light is absorbed and converted into heat energy by the MXene sheets, so the temperature of the top surface of the sponge rises rapidly (Fig. 4b)11,14,43. Under one-sun illumination exposed to air, the central temperature of the upper surface of the rGMX20 sponge increases rapidly from 22 to 55 °C within 10 s. The increase in MXene content improves the sponge’s photothermal conversion efficiency, and the central temperature of the upper surface of the rGMX5 sponge can reach 72 °C within 10 s (ΔT = 50 °C, Fig. 4c). Moreover, a fast photothermal response rate is also of great importance for the evaporators12,21,27,28. Therefore, the light (one-sun illumination) is turned off after 20 s of illumination (Supplementary Fig. 7b). After 10 s of illumination, the temperature distribution of the sponge tends to be stable and reaches a dynamic balance. The maximum temperature at the lateral surface of the sponges is lower than the temperature at the center of the top surface of the sponges. The temperature gradually decreases from top to bottom (Supplementary Fig. 7c). After the light is turned off, the sponge returns to its initial state after 20 s. In contrast, the temperature gradient from top to bottom is maintained in this process as the radial thermal conductivity of the sponge is far better than that of the axial direction.

Fig. 4: Photothermal performance of rGMXn sponges under one-sun illumination.
figure 4

a The light absorption spectra of rGMXn sponges. b Infrared images and c temperature of the top surface of rGMXn sponges under one-sun illumination exposed in the air for 30 s. d Schematic of the photothermal mechanism of rGMXn sponges. e Infrared images and f temperature of the top surface of rGMXn sponges under one-sun illumination floating on the water for 5 min. g Mass loss and h evaporation rate and efficiency of rGMXn sponges under one-sun illumination.

Full size image

The rGMXn sponges can float on the water and be used as a free-standing photothermal water evaporation device due to their high hydrophilicity and low density, thus guaranteeing adequate water supply without excess supporting materials or external energy input (Fig. 4d, Supplementary Fig. 8a). The rGMXn-sponge-based evaporators are tested under one-sun illumination (1.0 kW m−2) in the laboratory. The central temperature of the top surface of the rGMX20 sponge immersed in water has increased by 10 °C when exposed to one-sun illumination for 5 min, while the rGMX5 sponge with more MXene content has increased by 19 °C (Fig. 4e, f). Compared to exposure to the air (Fig. 4b), the relatively lower temperature is caused by the violent evaporation of water, carrying away most of the heat6,11,17,29. The water evaporation rates of the rGMXn sponges under one-sun illumination are 1.75 kg m−2 h−1 (rGMX20), 1.95 kg m−2 h−1 (rGMX10), and 2.35 kg m−2 h−1 (rGMX5), respectively (Fig. 4g, Supplementary Fig. 8b). Due to etching, the MXene nanosheets form a large number of hydrogen bonds with water by the abundant hydrophilic functional groups (\(-\)OH, \(-\)F, etc.) on the surface, which significantly improve the hydrophilicity of the sponges and reduce the enthalpy of evaporation of water on the sponge surface (Supplementary Fig. 9a)18,25. Raman spectra are used to investigate the hydratable interaction of the rGMXn sponges (Supplementary Fig. 9b). The region of O–H stretching in the Raman spectra represents the form of hydrogen bonds: the peaks at 3514 and 3630 cm−1 correspond to intermediate water (IW) with weak or non-hydrogen bonds, and the peaks at 3233 and 3401 cm−1 could be associated with free water (FW) with four hydrogen bonds (two lone electron pairs and two protons)17,52. A lower proportion of FW predicates a lower energy demand for water vaporization. The IW/FW value of water in the rGMX5 sponge reaches ca. 0.83, much larger than that of the rGMX10 sponge (0.74), rGMX20 sponge (0.69), and bulk water (0.29), indicating easier evaporation of water in the rGMX5 sponge (Supplementary Fig. 9c)17. The hydration state is examined by the differential scanning calorimetry (DSC) method, and the theoretical enthalpy of water evaporation in the rGMXn sponges is calculated through the DSC results (Supplementary Fig. 9d)7,18. The photothermal evaporation efficiency of sponges is calculated according to the theoretical enthalpy (Fig. 4h). The rGMX5 sponge with an evaporation rate of 2.35 kg m−1 h−1 under one-sun illumination possesses an evaporation efficiency of 127% with the MXene content of only ~4 mg ml−1 (Supplementary Table 1), much superior to previously reported photothermal evaporators (Supplementary Fig. 5b)11,24,30,42.

The continuous and effective evaporation efficiency of the rGMXn sponges benefits from the excellent integrated water–thermal management and the mechanism can be explained as follows: 1. MXene and rGO nanosheets within the rGMXn sponges possess high theoretical photothermal conversion efficiency in the whole frequency band of sunlight, ensuring sufficient heat energy for the evaporation of water in the sponges. Meanwhile, the MXene nanosheets vertically aligned on the rGO framework, dramatically enlarging the effective surface area of MXenes, further improving the photothermal conversion efficiency of sponges7,11,14. Besides, the hierarchical microstructure makes the incident sunlight multi-reflect among the MXene nanosheets and the rGO framework, enhancing sunlight absorption31,50. 2. The water management is accomplished by the sponges’ axially oriented vessels, which offer fast channels for water transfer from the bottom to the top (Supplementary Fig. 10)11,18,29. 3. The heat regulation is achieved by the low anisotropic thermal conductivity of the sponges: the relatively better radial thermal conductivity makes the temperature of the top surface of the sponges tend to be homogenized, achieving a larger area to participate in effective evaporation; the relatively slower thermal diffusion along the axial direction reduces heat loss for heating unnecessary water not participating in evaporation (Supplementary Fig. 11)6. 4. The abundant hydrophilic functional groups (\(-\) F,\(-\) OH, etc.) on the sponge surface significantly reduce the surface tension of water, reducing the evaporation enthalpy of water in the sponges17,52,53.

Evaporation performance of the sponges in actual environments

As the natural sunlight usually is lower than 1-sun, evaporation performance under weak irradiance is also of practical importance. Therefore, the rGMXn sponges are tested under weak sunlight of 0.3, 0.5, and 0.7-sun irradiance, respectively (Supplementary Fig. 12). The increase of MXene content dramatically improves the performance of the sponges under weak light. The rGMX5 sponge possesses an evaporation rate of 2.1 kg m−2 h−1 (0.7-sun), 1.6 kg m−2 h−1 (0.5-sun), and 1.2 kg m−2 h−1 (0.3-sun), respectively, which are comparable to the previously reported state-of-the-art evaporators4,11,43. To further test the stability of the sponge under actual working conditions, especially under weak sunlight, the rGMX5-sponge-based evaporator has been placed on a windowsill in the laboratory of Harbin Institute of Technology (HIT, Harbin, China). The system’s mass change and solar intensity have been carefully recorded from 7:00 to 17:00 to evaluate the practical performance of the rGMX5 sponges quantitatively (Fig. 5a). The system can yield 10.7 kg m−2 purified water and possess an average evaporation rate of ~1 kg m−2 h−1 in a single day. Due to the in-situ synthesis and etching, the exposure of MXene in the water–oxygen environment is avoided during fabrication, thereby extending its working life. The rGMX5 sponge exhibits a durable photothermal evaporation performance after sustained operation for ten days (Fig. 5b, Supplementary Fig. 13a). Furthermore, the rGMXn sponge shows a long stable working life and has a photothermal evaporation efficiency (2.26 kg m−2 h−1) after being soaked in water for 30 days (Supplementary Figs. 13b and 14).

Fig. 5: Photothermal water evaporation performance of rGMX5 sponges.
figure 5

a Synchronous solar intensity and water evaporation rates of rGMX5 over time under a real sky. b Mass loss and average mass loss rate of rGMX5 under a real sky for multi-days. Water evaporation rates of rGMX5 sponge in c a wide range of salinity (inset images show no salt deposition on the surface of rGMX5 sponge) and d a wide range of pH value. e Resistance of the water samples: I. Artificial seawater (salinity of 100‰); II. Water obtained from melting snow containing snow melting agent; III. Purified water of sample II (inset image shows the resistance test of water).

Full size image

In actual scenarios, the working water environment of the evaporators is unpredictable. Therefore, the sponges need to be able to adapt to various extreme environments. The rGMXn sponges possess strong mechanical stability and excellent hydrophily, ensuring they can be used as free-standing solar-driven evaporators floating on water. As expected, the rGMX5 sponges can maintain high evaporation efficiency under high salt conditions (Fig. 5c), with an evaporation rate of 1.75 kg m−2 h−1 in artificial seawater (300‰ salinity). Furthermore, the rGMX5 sponges can bear acid rain (pH < 4) and cleaning agents (pH > 12) with an evaporation rate higher than 2 kg m−2 h−1 (Fig. 5d). Water obtained from melting snow containing snow-melting agents has been used to further prove the practicability of the rGMXn sponges (Fig. 5e). After purification, the electrical resistance values obtained by a multimeter at a constant distance between electrodes increase from 120 kΩ to 1.27 MΩ, indicating improved water quality11,14,18,25. The excellent salt crystallization resistance of rGMXn sponges derives from two main factors. On the one hand, the MXene sheets are vertically aligned on the surface of the rGO framework, which greatly increases the MXene’s interlayer spacing54, accommodating more ions through intercalation55,56. On the other hand, the abundant directional channels enable the sponges with good ion diffusion/transportation kinetics57,58 and diffuse concentrated salt down into the water14.

Conclusions

In summary, passive water-thermal integrated management is achieved by anisotropic designing of the microstructure of the MXene-rGO sponges. The MXene nanosheets in the rGMXn sponges are in situ synthesized on the surface of the rGO framework, guaranteeing the sponges with high photothermal conversion efficiency in the whole solar wavelength. Furthermore, the sponges’ axial-oriented vessels provide fast water transmission channels for continuous photothermal water evaporation at the interface. The sponges’ radial thermal conductivity (0.268 W m−1 K−1) is much greater than that of the axial one (0.025 W m−1 K−1), reducing the heat loss, thus achieving excellent water and thermal integrated management. The obtained rGMXn sponges possess strong mechanical stability, low density, and excellent hydrophily and can be used as free-standing and durable steam generators floating on the water. Furthermore, the rGMXn sponges have excellent solar energy utilization efficiency, especially in weak natural sunlight. At a weak solar irradiance of 0.5-sun, the rGMX5 sponge has achieved an evaporation rate of 1.6 kg m−2 h−1 with an energy efficiency of 85%, illustrating potential applications for solar-driven water purification and desalination in the high-latitude water-deficient areas (under weak sunlight). The in-situ synthesis method proposed in this work paves an avenue for the low-cost realization of water and thermal integrated management of sponge-based solar-driven evaporators with easy deployment and long working life.

Methods

Materials

Titanium powders (Ti, 10 μm) were supplied by Beijing Xing Rong Yuan Technology Co., LTD, China. Aluminite powders (Al, purity of 99.9%, 300 mesh) were provided by Henan Yuanyang Powder Technology Co., Ltd., China. Sodium chloride (NaCl) and potassium chloride (KCl) were supplied by Tianli Chemical Reagent Co., Ltd., China. All the materials and chemicals were used without any treatment.

Synthesis of rGMXn sponges

Take the rGMX5 sponge as an example. 25 ml salt solution with 2.922 g NaCl and 3.95 g KCl was added dropwise to 100 ml concentrated GO slurry (6.25 mg ml−1). Afterward, 0.09 g Al powders and 0.638 g Ti powders were added to the colloidal suspension, and the GO/Al/Ti slurry precursor of rGMAX5 was obtained. The precursor rGMAX5 slurry was then dropped into a Teflon mold for directional freezing with liquid nitrogen. After demolding, the frozen samples were kept in a freeze dryer for freeze-drying. Subsequently, the sponges were set in the tube furnace for heat treatment. With the protection of Ar gas, the tube furnace was heated to 500 °C at the rate of 3 °C min−1 and kept for 1.5 h, then heated to 800 °C at the rate of 2 °C min−1 and kept for 3 h, and then heated into 1100 °C with the rate of 3 °C min−1 and kept for 1 h. After that, the tube furnace was cooled down to room temperature naturally. Thereby, the rGMAXn sponges with residual salts within were obtained. After that, the as-prepared rGMAXn sponges were etched with an acid solution (10 wt% HF and 10 wt% HCl) at 40 °C for 10 h. Thereby, the sponges were soaked in deionized water (DI) to remove acid, and then the rGMXn sponges were obtained. The other rGMXn sponges are prepared with the same method (Supplementary Table 2), except the amount of metal powder is different, which is determined by the molar ratio of Ti, Al, and C (4:1: n).

Characterization

Scanning electron microscope (FIB-SEM; TESCAN AMBER, Brno-Czech Republic) with an energy dispersive spectrometer (EDS) detector was employed to analyze the microstructure and elements distribution of obtained samples. X-ray diffraction (Bruker D2 Phaser, Germany) equipment with Cu Kα radiation was used to measure the phase compositions of obtained samples ranging from 10° to 90° at a scan rate of 5° min−1. XPS (ESCALAB 250, USA) and Raman (inVia-Reflex, UK) were carried out to record the spectra data. Mechanical tests were carried out by an instrument (Instron 5944) equipped with two flat-surface compression stages and a 2000 N load cell. The anisotropic thermal conductivities of the sponges were measured using an LFA467 (NETZSCH, Germany). The sunlight absorption spectra of the sponges were measured using a UV–vis–NIR spectrophotometer (UV3600i Plus, Shimadzu, Japan). The enthalpy of evaporation of water in the sponges was calculated based on differential scanning calorimeter (DSC) data collected by a thermogravimetry-differential scanning calorimetry system (NETZSCH STA 449F3, Germany) in a nitrogen environment.

Solar-driven vapor generation experiments

The water to be purified was held in an acrylic petri dish, and the rGMXn sponges were self-floated on the water surface. The illumination was applied using a light source (CEL-HXUV300), and the solar intensity was measured by a solar power meter (CEL-NP2000-2(10)A, Beijing Zhongjiao Jinyuan Technology Co., Ltd). The mass change of water was recorded every hour by an electronic mass balance (at an accuracy of 0.1 mg), while the temperature change was monitored by a Thermal Imager (VarioCAM HD head, Germany). The room temperature is 20 °C with humidity at 40% in the laboratory. The actual solar-driven evaporation of sponges was carried out indoors under natural sunlight for ten consecutive days from 7 to 16 March 2023 in Harbin.