1 Introduction

The global freshwater resource scarcity has triggered a drinking water crisis, which is related to human survival [1,2,3,4,5]. Seawater desalination is considered to be one of the most potential solutions [6, 7]. Solar-driven water evaporation technology has increasingly been recognized as a promising strategy for seawater desalination compared with traditional methods [8,9,10]. To this end, the development of photothermal conversion materials with enhanced light absorption is the prerequisite for high-efficiency solar water evaporator.

To date, various photothermal materials have been widely used for seawater evaporation, including inorganic semiconductor materials [11, 12], carbon-based materials [13, 14], organic polymers [15, 16], surface plasmon resonance metals [17, 18], etc. Thereinto, two-dimensional MXene nanosheets with large surface areas and broad photo-absorption ability have been widely employed in photocatalysis and photothermal conversion as light absorbers [19, 20]. The photothermal conversion mechanism of MXene is as follows: electrons-holes pairs could be generated in MXene when they were illuminated by photons with energy larger than its band gap potential; excited electrons tend to return to the ground state to bond with holes, converting radiant energy into thermal energy at the same time [21, 22]. However, MXene is prone to agglutination, which not only limits its photothermal conversion capacity but also results in fewer interlayer transport channels for water [23]. This is extremely detrimental to seawater desalination, so it is still urgent to search for a high-efficiency photothermal conversion material.

Studies have shown that the self-assembly of Ni chain and MXene (MXene/Ni) by electrostatic action can not only effectively solve the agglomeration phenomenon of MXene by avoiding the stacking of MXene lamellar, but also further improve the photothermal conversion performance of materials [24]. As a typical plasmon resonance metal, Ni has broadband absorption, the maximum absorption rate of sunlight can be up to 96% [25,26,27]. Increasing the Ni chains size could further enhance its absorption of solar energy [28, 29], thus improving the photothermal conversion efficiency [14]. However, the hydrophobicity of Ni chains limits the application of MXene/Ni in seawater evaporation. Incorporating cellulose nanofiber (CNF) can effectively improve the hydrophilicity of MXene/Ni because of the abundant –OH on CNF surface [30, 31]. Meanwhile, the three-dimensional (3D) skeleton structure of CNF is conducive to the enhancement of water transport channels in CNF/MXene/Ni chain system.

In this study, CNF/MXene/Ni chain (CMN) aerogel was successfully constructed by electrostatic self-assembly and hydrogen bond crosslinking. Due to the enhanced light absorption and photothermal conversion of Ni chain and MXene, the hydrophilicity of CNF and the abundant porous structure in CMN aerogel, the water evaporation rate of CMN aerogel is significantly improved. CMN aerogel could implement a maximum evaporation rate as high as 1.85 kg m−2 h−1 in pure water, while 1.81 kg m−2 h−1 in seawater under one sunlight intensity. Furthermore, the concentration of Ca2+, K+, Na+ and Mg2+ in seawater could be reduced by CMN to the range required by WHO and the USA for drinking water quality. Some heavy metal ions and organic pollutants could also be effectively rejected and the rejection rate of Rhodamine B and other pollutants is above 99.69%. Therefore, this study provides new methods and ideas for solving the water resources crisis and the development and utilization of solar energy.

2 Experimental

2.1 Sample preparation

2.1.1 Preparation of MXene

In the first step, LiF (1 g) and deionized water (5 mL) were added into 15 mL concentrated hydrochloric acid, followed by adding 1 g Ti3AlC2 (MAX) under stirring. After the reaction at 40 °C for 24 h in a water bath, the obtained solution was centrifugally washed at 3500 rpm for several times until neutral. The washed solution was transferred to a conical flask, sonicated for 3 h in an ice water bath. MXene dispersion solution was collected after centrifugation (5000 rpm for 20 min).

2.1.2 Preparation of P-Ni chain

0.594 g of nickel chloride hexahydrate (NiCl2·6H2O) was dissolved in 50 mL ethylene glycol, and then, 1 g of polyvinylpyrrolidone (PVP) and 0.4 g of sodium hydroxide (NaOH) were added into above solution, ultrasonically dispersed for 30 min in an ice water bath. Then, 1.5 mL of hydrazine hydrate (N2H4·H2O) was dropped into the mixture and continued to the ultrasound for another 10 min. Finally, under the parallel magnetic field, the mixture was bathed in water at 80 °C for 1 h, and Ni chain was obtained after washing centrifugally. Poly diallyl dimethyl ammonium chloride (PDDA 1 wt%) modified Ni chain (P-Ni chain) was achieved by employing PDDA into the aforementioned Ni chain solution with 30 min ultrasound.

2.1.3 Preparation of CMN aerogel

The above-prepared MXene dispersed solution (8.5 mL 10 mg L−1) and P-Ni chain (0, 2.5, 5, 7.5, 10, and 12.5 mg) were added into the CNF solution (1.5 g, 1 wt%) and stirred vigorously for 2 h to obtain the suspension. The mixture was then transferred to a mold, frozen in a refrigerator for 24 h, followed by vacuum freeze-dried (− 78 °C, 20 Pa) for 48 h to get CNF/MXene/Ni chain (CMN) aerogel. The samples with different P-Ni chain contents were synthesized, labeled as CMNX (X represents the incorporated mass of P-Ni chain (mg), X = 0, 2.5, 5, 7.5, 10 and 12.5). Figure 1 shows the preparation process of CMN aerogel.

Fig. 1
figure 1

Schematic synthesis process of CMN aerogel

Full size image

2.2 Solar desalination test

The evaporation test was conducted at ~ 15 °C and humidity of ~ 45%. Xenon lamp equipped with AM1.5G filter (1 kW m−2, Beijing CEJ Tech. Co., Ltd., Beijing, China) was used as the sunlight simulation light source. In the experiment, the CMN aerogel was cut into a square with an area of 9 cm2, and the initial water volume was 100 mL. The CMN aerogel deposited foam board (6 cm in diameter) served as light absorber, which was placed on the water surface. The hydrophilic cotton with capillary action was closely contacted with the foam board as a water transportation channel. Under light irradiation, the weight of the balance was recorded every 5 min and lasted for 1 h. The seawater desalination test was conducted with simulated seawater which is configured according to Table S1. All experiments were repeated three times. The water vapor evaporation rate was calculated as follows [32]:

$$\begin{array}{*{20}c} {{\text{Evaporation }}\;\left( {{\text{kg m}}^{ - 2} {\text{ h}}^{ - 1} } \right) = \frac{{m_{t} }}{S \times t}} \\ \end{array} ,$$
(1)

where m (kg) represents the change of water mass, S (m2) represents the actual area of the sample, and t (h) represents the evaporation time of water.

The evaporation efficiency was calculated according to Eqs. (2) and (3) [33, 34]:

$$\begin{array}{*{20}c} {\eta = \frac{{m\left( {H_{{{\text{LV}}}} + C \times \Delta T} \right)}}{{P_{{\dot{i}n}} }}} \\ \end{array} ,$$
(2)
$$\begin{array}{*{20}c} {H_{{{\text{LV}}}} = 1.91846 \times 10^{6} \left[ {T/\left( {T - 33.91} \right)} \right]^{2} } \\ \end{array} .$$
(3)

In the formula, η (%) represents the evaporation efficiency of the sample, m (kg m−2 h−1) is the mass change rate of evaporated water, C (4.18 kJ kg−1 K−1) is the specific heat capacity of water, ΔT (K) is the temperature change on the surface of the sample, and Pin (kJ m−2 h−1) is the optical power density, HLV (kJ kg−1) is the enthalpy of water vapor phase transformation, and T (K) is the evaporation temperature.

In the evaporation process, the surface temperature of the sample is tested and recorded every 30 s by the near-infrared imager (HIKMICRO, H21); in order to test the rejection rate of organic pollutants, the Ultraviolet/Visible/Near-infrared (UV–Vis–NIR) spectrophotometer (Beijing Purkinje TU-1900) was used to select the corresponding maximum absorption peak of pollutants for concentration monitoring, and the corresponding concentration was obtained according to Fig S1.

For the seawater desalination test, the concentrations of Na+, Ca2+, K+ and Mg2+ in the simulated seawater were 10,780, 410, 380 and 1310 mg L−1, respectively. The pH of simulated seawater was adjusted by HCl and NaOH. 10 mg L−1 heavy metal solution and 30 mg L−1 organic dye contaminant with simulated seawater as solvent were employed in the pollutant rejection experiments. The rejection rates for metal cations, heavy metal ions, and organic pollutants in seawater were calculated according to Eq. (4):

$$\begin{array}{*{20}c} {{\text{Rejection}}\;{\text{rates}} = \frac{{C_{1} - C_{2} }}{{C_{1} }} \times 100\% .} \\ \end{array}$$
(4)

In the formula, C1 represents the initial concentration of pollutants, and C2 represents the concentration of pollutants in the condensate.

3 Results and discussion

3.1 Characterization of CMNx

The microstructure of P-Ni chain, MXene, and CMN10 aerogel was observed by scanning electron microscope (SEM) and transmission electron microscope (TEM) (Fig. 2). The P-Ni chain was formed by directional assembly of Ni microspheres, with a diameter of about 150 nm and a length-to-diameter ratio of above 20 (Fig. 2a). The facilitated light absorption by the increased aspect ratio was favorable for improving the photothermal conversion performance of CMN [29]. As shown in Fig. 2b and S2, MAX after etching presented a typical accordion package, and monolayer MXene nanosheets appeared after ultrasonic stripping. The X-ray diffractometer (XRD) patterns of MAX and MXene in Fig. S3 demonstrated the successful synthesis of MXene (the disappearance of the diffraction peak at 38.9° means the removal of Al atomic layer) [35]. The SEM images of CMN10 aerogel cross section displayed that the CMN10 aerogel maintained the 3D porous honeycomb structure of CNF (Fig. 2c); P-Ni chain was anchored on the MXene lamella by electrostatic force, and MXene nanosheets were attached to the pore walls of CMN10 (Fig. 2d). This manifested the CMN10 aerogel which was prepared successfully. As shown in Fig. 2e–h, the energy dispersion spectrum (EDS) mapping further verified the successful anchoring of P-Ni chain onto the MXene lamellae and the synthesis of CMN aerogel. Studies have shown that the rich pore structure in a sample can produce multiple reflections of sunlight, thus enhancing the absorption of sunlight [36]. In addition, the rich pore structure also facilitates the rapid escape of water vapor generated, thus increasing the evaporation rate of water vapor of material [37]. The anchoring of P-Ni chain can effectively avoid the stacking of MXene layers and promote the light absorption of CMN. Therefore, CMN aerogel have great potential in solar desalination.

Fig. 2
figure 2

a SEM image of P-Ni chain; b TEM image of MXene lamellar; SEM image of c–e CMN10 aerogel; f–h EDS element mapping results for Ni, Ti and C

Full size image

The theoretical feasibility of successful preparation of CMN was demonstrated by Zeta potential (Fig. 3a) and the Fourier transform infrared (FTIR) (Fig. 3b). As exhibited in Fig. 3a, the surface charge of MXene was − 32.6 mV due to the abundant –OH on the surface of it, while + 44.1 mV for P-Ni chain because of the modification of PDDA. Electrostatic self-assembly easily occurs between MXene and P-Ni chain with opposite charges [38]. The abundant –OH on the surface of CNF, MXene, and CMN10 were further certified by FTIR spectra (Fig. 3b). In the spectra of CNF and MXene, peaks at 3445 cm−1 and 1660 cm−1 were observed as –OH stretching vibrations, while the peak at 1050 cm−1 was C–O stretching vibration peak. The characteristic peaks of CNF and MXene can be found in the FTIR spectra of CMN aerogels, which indicates the successful complexation of MXene with CNF [39]. It was observed that the peak position of C–O shifted from 1050 cm−1 (in MXene and CNF) to 1040 cm−1 (CMN10), which was mainly due to the formation of a large number of intermolecular hydrogen bonds between CNF and MXene [40, 41]. The hydrogen bond interactions formed by –OH crosslinking promoted the close connection between CNF and MXene. Furthermore, the electrostatic interaction between MXene and P-Ni chain as well as the hydrogen bonding effect among MXene and CNF is illustrated in Fig. 3c.

Fig. 3
figure 3

a Zeta potential of MXene nanosheet, CNF and P-Ni chain; b FTIR spectra of CMN10, MXene nanosheet and CNF; c electrostatic interaction and hydrogen bonding effect among MXene, CNF and P-Ni chain; d XRD patterns of CMN10, MXene nanosheet, CNF and P-Ni chain; e the XPS spectra and f Ni 2p high-resolution spectra of CMN10

Full size image

Figure 3d shows the XRD pattern of CMN10, P-Ni chain, CNF and MXene. Compared with MXene, the (002) peak of MXene in CMN10 shifted from 6.18° to 5.8°, which is due to the electrostatic force and hydrogen bond among P-Ni chain, MXene, and CNF, leading to the increase in layer spacing in MXene [42]. In addition, peaks at 44.5°, 51.6° and 76.4° in the XRD pattern of CMN10 corresponded to (111), (200) and (220) crystal planes of Ni, respectively [43], which means that the P-Ni chain has been successfully reassembled in CMN10 by electrostatic self-assembly. The composition and valence structure of the elements in CMN10 were explained by X-ray photoelectron spectroscopy (XPS). As shown in Fig. 3e, the coexisting of Ni, C, O, Ti, and F elements in the survey spectrum (Fig. 3e) indicated the successful preparation of CMN10. In the high-resolution XPS spectrum of C 1 s (Fig. S4a), peaks at 288.58, 286.18, 284.58, 282.28, and 281.58 eV were corresponding to O–C=O, C–O, C–C, Ti–C–O, and C–Ti, respectively [44]. As observed in Fig. S4b, the Ti–C 2p3/2, Ti(II) 2p3/2, TixOy 2p3/2, TiO2 2p3/2, Ti–C 2p1/2, Ti(II) 2p1/2, TixOy 2p1/2, TiO2 2p1/2 of MXene located at 454.68, 455.28, 456.08, 456.98, 458.08, 460.68, 461.68, and 463.78 eV, respectively [45, 46], indicating that MXene was successfully doped in CMN10 and maintains the intrinsic state. The high-resolution spectrum of Ni 2p in Fig. 3f displayed the peaks at 855.38 and 873.08 eV, which were assigned to Ni 2p1/2 and Ni 2p3/2, respectively [47]. Figure S5 is the TGA curve of CMN10 that demonstrated the excellent thermal stability [48]. Only about 10% mass loss was shown when the temperature was as high as 800 °C.

The light absorption and photothermal conversion of materials are critical to obtaining excellent water evaporation performance [49]. Infrared imaging was performed to detect the surface temperature change of CMNX (X = 0, 2.5, 5, 7.5, 10, and 12.5) under one sun illumination (Fig. 4a). A darker color indicates a higher temperature. As shown in Fig. 4a, the color of CMN10 was much darker than other CMNX (X = 0, 2.5, 5, 7.5, and 12.5). The surface temperature of CMN10 was as high as 58.5 °C, while only 45.3 °C can be detected on the surface of CMN0. Nevertheless, the surface temperature of CMN increased first and then decreased with the increase in P-Ni chain content. This could be for two reasons: (1) excess P-Ni chain leads to their possible aggregation on the aerogel surface, which increased the sunlight reflection; (2) the heat conduction of CMN was faster when the thermal conductivity increased caused by excess P-Ni chain [50]. The light absorption ability of CMN0 and CMN10 is further tested by UV–Vis NIR spectrophotometer. It can be seen from Fig. 4b that CMN10 exhibited full spectrum absorption within the range of 200–2500 nm. The absorption of sunlight by CMN10 is up to 95%, while that of CMN0 is only about 61%. Based on Fig. 4a–b, it can be concluded that the light absorption capacity of CMN is proportional to its photothermal conversion capacity [19].

Fig. 4
figure 4

a The infrared image of CMNx in sunlight; b optical absorption spectra of CMN10 and CMN0 in sunlight; c water contact angle of CMNx

Full size image

The hydrophilicity of the sample is another indicator to evaluate the desalination performance of it [51]. As shown in Fig. 4c, water contact angles of CMN0, CMN2.5, CMN5, CMN7.5, CMN10 and CMN12.5 were carried out to evaluate their hydrophilicity. All the samples were able to absorb the drops within 20 ms after the water drops off because of the abundant –O–H (Fig. S6). Moreover, the addition of P-Ni chain did not affect the hydrophilicity of the sample.

3.2 Solar-powered water evaporation performance of CMNX aerogels

The evaporation rate is the most intuitive indicator to judge the vapor evaporation performance of the samples; therefore, pure water was first used as the evaporation source to test the vapor evaporation performance of CMN aerogel [52]. The water evaporation device is shown in Fig. S7a. The water vapor can be obviously seen on the top of the device in dark (Fig. S7b). Figure 5a reflects the mass changes of water over various CMN under light irradiation. With the irradiation time increased, the mass of all samples decreased, with CMN10 being the most pronounced. The evaporation rate of CMN10 reached the highest of 1.85 kg m−2 h−1 (Fig. 5b). However, the evaporation rate of CMN12.5 was much lower than CMN10, which was consistent with the light absorption and photothermal conversion properties of the sample (Fig. 4a–b). The reasons for this phenomenon are as follows: (1) the excess Ni chains form a pile-up and lead to an increase in the reflectance of CMN12.5 to sunlight, which reduces the photothermal conversion efficiency of CMN12.5; (2) the excess addition of Ni chains leads to CMN12.5’s thermal conductivity will increase, and the heat generated by the photothermal reaction will be rapidly conducted and dissipated, and cannot be used to heat water and cause a decrease in the evaporation rate of water vapor [50]. The variation trend of the maximum surface temperature of the sample during evaporation (Fig. 5c) is similar to that of the evaporation rate (Fig. 5a, b), and the surface temperature of CMN10 is the highest, reaching 40.5 °C. Compared with CMNX (X > 0), the surface temperature of CMN0 rose slowly, reaching 37.4 °C in about 7 min, which indicated the favorable role of P-Ni chain in promoting the light absorption and photothermal conversion of sample.

Fig. 5
figure 5

a The mass changes of water over time; b evaporation rates and surface temperatures; c temperature rise curves; d solar evaporation efficiency of CMN0, CMN2.5, CMN5, CMN7.5, CMN10 and CMN12.5

Full size image

Solar evaporation efficiency is also one of key paramenters to measure the performance of water vapor evaporation of materials, which directly reflects the utilization degree of sunlight [53]. It can be observed from Fig. 5d that the solar evaporation efficiency was also positively correlated with the water vapor evaporation rate. The solar evaporation efficiency of CMN10 reached up to 96.04%, much higher than that of CMN0. The stability test showed the excellent cyclic stability in evaporation performance of CMN10 (Fig. S8). This could be attributed to the synergy of electrostatic and hydrogen bond interaction among CNF, MXene, and P-Ni chain. Good stability is significantly critical in practical application of the sample.

CMN10 was screened for the following seawater desalination experiment based on the above optimal water evaporation. Under one sun irradiation (1 kW m−2), pure water, simulated seawater, 7 wt%, 10 wt%, and 20 wt% NaCl solution were, respectively, used for desalination performance test. As shown in Fig. 6a–b, salt concentration has little effect on its evaporation rate, even in solution with 20 wt% NaCl concentration, the evaporation rate of CMN10 can still reach 1.72 kg m−2 h−1, which is only reduced by 7.03%. This also proved that CMN10 has a high salt tolerance. The cycle stability of CMN10 in seawater was measured with simulated seawater. As demonstrated in Fig. 6c, the water evaporation rate of CMN10 stabilized at 1.71 kg m−2 h−1 after 10 cycles. The above experiments showed that CMN aerogel has excellent cyclic stability and salt tolerance and CMN aerogel is a potential candidate for seawater evaporation in real environments.

Fig. 6
figure 6

CMN10 as an evaporator, a mass change curve and b evaporation rates of seawater from pure water, simulated seawater, 7 wt%, 10 wt%, 20 wt% NaCl solutions; c cycle stability of simulated seawater; d water mass change curves of simulated seawater at different light intensities; e water evaporation rates of simulated seawater at different pH values; f comparison of concentrations of major cations in simulated seawater before and after desalination

Full size image

The water vapor evaporation rate of CMN10 under different light intensities also performed (Figs. 6d, S9). As expected, the water evaporation rate of CMN10 aggrandized with light intensity increases, and it is 1.81, 3.41, and 4.95 kg m−2 h−1 under 1 sun, 2 sun and 3 sun irradiations, respectively. Obviously, there is no linear correlation between the evaporation rate and the light intensity. This could be ascribed to the increasing temperature difference between CMN10 surface and the environment with the increase in the light intensity; meanwhile, the thermal convection and radiation increase accordingly, resulting in energy loss.

Considering the complexity of real seawater, the evaporation rate of CMN10 in simulated seawater with different pH was further evaluated. As shown in Figs. 6e and S10, the actual water evaporation rate of CMN10 can still reach up to 1.74 and 1.73 kg m−2 h−1 in simulated seawater with pH 2 and 12, respectively. All the above evidence implied the stability and the practical application potential of CMN aerogel.

The rejection rate of metal ions is an important index of seawater desalination. Fig. S11 shows the experimental device for collecting the cooling water after the evaporation of simulated seawater. The ion concentration in cooling water was measured by inductively coupled plasma mass spectrometry (ICP-MS) to evaluate the seawater purification rate of CMN10. As observed in Fig. 6f, Na+ concentration decreased from 10,780 to 1.05 mg L−1, which is far below the WHO standard of drinking water quality (200 mg L−1) and the stricter requirements of the US Environmental Protection Agency (EPA) drinking water quality standards (20 mg L−1). The rejection rate of Na+ over CMN10 reached an astonishing 99.99%. As anticipated, the concentrations of Ca2+, K+, and Mg2+ in the collected cooling water were 1.29, 0.53, and 0.21 mg L−1, respectively, also much lower than the WHO standard of drinking water quality, and the corresponding ions rejection rates reached 99.68%, 99.86% and 99.98%, respectively.

With the aggravation of environmental pollution, seawater desalination should pay more attention to the content of heavy metal ions and organic pollutants in cooling water. Therefore, the rejection ability of CMN10 was tested with Pd2+, Cd2+, and Cr2+ as model heavy metal ions. As displayed in Fig. 7a, the concentration of heavy metals in cooling water has significantly decreased, the concentration of Pd2+, Cd2+, and Cr2+ in cooling water is, respectively, 0.0007, 0.0013, 0.0005 mg L−1 and far below the drinking water requirements of WHO. The rejection rates of Pd2+, Cd2+, and Cr2+ by CMN10 are all above 99.98%. In addition, Rhodamine B, Congo red and Methylene blue were used as model organic pollution sources to test the rejection ability of CMN10 for organic pollutants in seawater. After CMN10 purification, the concentrations of Rhodamine B, Congo red and Methylene blue in cooling water were reduced to 0.0097, 0.0917 and 0.0098 mg L−1, respectively, and the rejection rate reached more than 99.69% (Fig. 7b, c). The excellent rejection ability of dyes can be confirmed by the color differences in visual optical photographs (Fig. 7d–f). All these properties indicated that CMN materials have great potential in seawater desalination.

Fig. 7
figure 7

CMN10 as evaporator, a concentration changes and rejection rates of heavy metal ions in simulated seawater; b concentration change before and after treatment, c rejection rate, and d–f digital photographs of organic pollutants

Full size image

4 Conclusion

To sum up, CMN aerogel was successfully constructed by electrostatic self-assembly and hydrogen bond crosslinking. The rich pore structure provided by CNF not only facilitates the rapid escape of water vapor generated, but also produces multiple reflections of sunlight, thus enhancing the absorption of sunlight. Furthermore, the anchoring of P-Ni chain can effectively avoid the stacking of MXene layers and produce synergistic effect with MXene in the enhancement of the light absorption by CMN. CMN10 displayed high evaporation rates of 1.85 and 1.81 kg m−2 h−1 in pure water and simulated seawater, respectively. The rejection rate of metal ions (such as Na+, Ca2+, K+, and Mg2+; Pd2+, Cd2+, and Cr2+) and organic pollutions (Rhodamine B, Congo red and Methylene blue) in seawater was extremely satisfactory, and the obtained cooling water fully met the requirements of American drinking water quality standards. Most importantly, CMN10 is quite stable under complicated conditions, such as different salt concentrations, light intensity, and pH, which is of great significance for the actual desalination. This work is believed to shed new light on designing highly efficient seawater evaporation and purification materials for clean water production. Meanwhile, these functional nanocomposites can be used for other applications such as electromagnetic interference (EMI) shielding [54,55,56,57,58], energy usage [59,60,61,62,63,64], electronics [65,66,67], and coating for protecting metals [68].