2.1 Materials

Hydrochloric acid was purchased from Sinopharm Chemical Reagent Co., Ltd. Lithium fluoride. FeCl₃⋅6H₂O, FeCl₂⋅4H₂O, Cetyltrimethylammonium bromide (CTAB) and NaOH (10 M) were purchased from Aladdin. Myristic acid(MA, AR, 99%)was purchased from Macklin. Ti₃AlC₂ (MAX, 98%, 400 mesh) powder was obtained from Macklin.

2.2 Preparation of MXene

Multilayer MXene was prepared by selectively etching Al layer of MAX.1 g LiF and 1 g Ti₃AlC₂ (MAX) were slowly added to 20 mL HCl (10 M) solution under vigorous stirring for 36 h at 35°C. The obtained product was dissolved in deionized water and centrifuged at 3600 rpm for several times until the supernatant was neutral. The precipitate was freeze-dried to obtain multilayer MXene powder.

2.3 Preparation of MXene@Fe₃O₄

MXene@Fe₃O₄ was prepared using an in situ chemical synthesis strategy. Typically, 150 mg MXene was added to a three-necked bottle with 100 mL deionized water. 87.5 mg FeCl₃·6H₂O and 80 mg surfactant CTAB were added into the reaction system and were vigorously stirred under N₂ atmosphere at 55℃ for 0.5 h. Subsequently, 32.2 mg FeCl₂·4H₂O and NaOH standard solution were added to the mixture to adjust PH to be higher than 10, and then were mechanically stirred for 4.5 h. Finally, the product was washed with deionized water to adjust to neutrality and then freeze-dried to obtain MXene@Fe₃O₄. In this study, MXene@Fe₃O₄ with different ratios of MXene and Fe₃O₄ (4:1, 4:2, 4:3 and 4:4) were prepared by adding different quality of FeCl₂ and FeCl3 using the same method.

2.4 Preparation of MXene@Fe₃O₄-MA composite PCMs and MXene@Fe₃O₄-MA-PU film

MXene@Fe₃O₄-MA composite PCMs were prepared using a physical vacuum infiltration method. First, 350 mg MA was added to the ethanol solution at 80°C until MA was completely melted. Then, MXene@Fe₃O₄ powers were slowly added into the MA-ethanol solution under vigorous stirring for 1 h. Finally, the obtained mixture underwent vacuum drying to obtain MXene@Fe₃O₄-MA composite PCMs at 80°C for 24 h.

MXene@Fe₃O₄-MA solution was mixed with waterborne elastic polyurethane, in which the loading content of MXene in PU was 1.5 wt%. The resulting mixture was uniformly dispersed under ultrasound and stirring. Subsequently, flexible MXene@Fe₃O₄-MA-PU films were prepared by coating the solution onto the glass substrate and drying at 80°C for 5 h.

2.5 Simulations

All of the simulated spectra and electromagnetic field distribution were obtained by solving three-dimensional Maxwell equations with the commercial COMSOL Multiphysics software, which is based on Finite Element Method (FEM). The periodic boundary condition was used to simulate the LSPR effect of MXene and MXene@Fe₃O₄ under incident light. The boundary conditions were set as follows: 1) the incident light illuminated MXene@Fe₃O₄ in a direction of 45°, 2) the refractive index of Fe₃O₄ and MXene was 2.34 + 0.11i and -10.75 + 11.28i, respectively, 3) the interlayer spacing of MXene was 150 nm, and the size of Fe₃O₄ nanoparticle was 100 nm.

2.6 Characterizations

The microscopic morphologies were observed with scanning electron microscope (SEM, SU8010, Japan Hitachi). The functional groups were evaluated by Fourier transform infrared spectrometer (FT-IR, Nicolet 6700, USA). X-ray Diffraction (XRD, M21X, Cu Ka radiation at 40 kV and 150 mA, λ¼0.15406 nm) was used to define the crystal structures. The pore structure was analyzed by nitrogen adsorption and desorption analyzer (Micromeritics ASAP 2420). Thermogravimetric analyzer (TGA, 449F3, Germany) was carried out to explore the thermal stability from 40 to 800℃ at a heating rate of 10℃/min under N₂ atmosphere. Thermal properties of pristine MA and MXene@Fe₃O4-MA composite PCMs were characterized by differential scanning calorimeter (DSC, DSC3, Mettler Toledo, Switzerland) under 50 mL/min N₂ flow at a heating/cooling rate of 10°C/min. The thermal conductivity of sample was measured via a flash method with a laser thermal conductivity analyzer (Netzsch LFA 427, Germany) at room temperature. The samples were pressed into disc-shaped with the diameter of 12.7 mm and the thickness of 1 mm. Then, the pressed standard samples were caught in the disc of analyzer to complete the test. Each sample was tested three times to ensure the accuracy and reliability of the results. The light absorption capacity was evaluated by ultraviolet-visible-near-infrared (UV-Vis-NIR) spectrometer (Agilent, Cary 5000). Vibrating sample magnetometer (VSM) (Quantum Design, USA) was used to test the magnetic properties at room temperature. The thermal response rates were recorded by an infrared thermal imager. The photothermal energy conversion and storage capacity was tested under simulated solar (CEAULIGHT, CEL-S500), and the temperature-time curves were measured by a digital data collector (R2100). An alternating current generator was used to provide an alternating magnetic field for magnetic-thermal conversion and storage, and the temperature-time curves were measured by an optical fiber sensor. The in situ observation of crystal evolution of composite PCMs and MA were performed on a polarizing optical microscopy (POM).

3.1 Structural analysis of MXene@Fe₃O₄

Lithium fluoride and hydrochloric acid solutions were used to etch the Al atomic layer in the MAX phase (Ti₃AlC₂), thus yielding accordion-like multilayer MXene (Ti₃C₂) (Figure S1C,D). As seen from XRD patterns, the most peaks of MAX disappeared in MXene, especially the characteristic peak of 38.8° (Figure S1A), indicating the successful preparation of MXene. MXene@Fe₃O₄ composites were prepared by in situ synthesizing Fe₃O₄ in MXene (Figure 1A). Free Fe³⁺ and Fe²⁺ ions were attached to the surface and interlayer of MXene through electrostatic adsorption due to a large number of F^– and OH^– ions on the surface of MXene, therefore, Fe₃O₄ nanoparticles were in situ synthesized under alkaline conditions. Under the assisted action of a large number of negative ions on the surface and the layers of MXene, metal sources of magnetic nanoparticles were firmly anchored on MXene during the in situ synthesis process, and Fe₃O₄ nanoparticles were uniformly dispersed on the outer surface of MXene and the interlayer space, which would effectively solve the unevenness problems of mechanical, magnetic and heat transfer caused by the agglomeration of magnetic nanoparticles. To further confirm the uniform distribution of Fe₃O₄ nanoparticles in MXene, the element mapping test of MXene@Fe₃O₄ was carried out. As shown in Figure S2A–D, C, Ti and Fe elements were uniformly distributed in MXene@Fe₃O₄, indicating the successful preparation of MXene and the uniform distribution of Fe₃O₄ nanoparticles in MXene.

Meanwhile, the average layer spacing of MXene was about 150 nm, and the average size of Fe₃O₄ nanoparticles was about 100 nm, which provides sufficient space condition for the in situ anchoring of Fe₃O₄ nanoparticles in the interlayer of MXene (Figure 1B–E and Figure S2E). When the ratio of MXene to Fe₃O₄ was 4:1, the content of Fe₃O₄ nanoparticles was too small to observed on the surface and interlayer of MXene. When the ratio increases to 4:2 and 4:3, Fe₃O₄ nanoparticles can be obviously observed and evenly distributed on the surface and interlayer of MXene. However, with the increase of Fe₃O₄ content, Fe₃O₄ nanoparticles tended to agglomerate between MXene layers. Overly agglomerated Fe₃O₄ nanoparticles in MXene reduce the light absorption ability of MXene, thus influencing the photothermal conversion efficiency of MXene-based composite PCMs.

According to XRD patterns of MXene@Fe₃O₄ (Figure S1B), the peaks located at 30.0°, 35.3°, 43.0°, 57.0° and 62.7° corresponded to the crystal planes of (220), (311), (400), (511), and (400) of the face-centered cubic Fe₃O₄, respectively, and the characteristic peak of MXene (002) was also included, indicating the successful preparation of MXene@Fe₃O₄. In addition, the (002) peak of MXene@Fe₃O₄ shifted to a low angle (Inset of Figure S1B) compared with MXene, because the in situ formation of Fe₃O₄ in MXene expanded the interlayer spacing of MXene. To characterize the pore structure of MXene@Fe₃O₄, N₂ adsorption-desorption isotherm was determined (Figure S3). The isotherm of 4:1 MXene@Fe₃O₄ was type IV with an obvious H3-type hysteresis loop, which was typical for a mesoporous material. The pore sizes of 4:1 MXene@Fe₃O₄ were mostly distributed between 2–15 nm, which clearly suggested the existence of mesoporous structure. Compared with macropores, mesopores can provide sufficient capillary force to prevent the leakage of liquid PCMs. Mesopores also can provide sufficient space for the free crystallization behavior of PCMs, which can efficiently store and release more heat, and provide mass and heat transfer channels. Therefore, our constructed mesoporous MXene@Fe₃O₄ provides the possibility to obtain MXene@Fe₃O₄-based composite PCMs with excellent thermal properties.

3.2 Thermal analysis of MXene@Fe₃O₄-MA composite PCMs

MXene@Fe₃O₄-MA composite PCMs were prepared via a vacuum impregnation method. Before loading, vacuum activation of MXene@Fe₃O₄ was performed to remove the adsorbate in the pores and open the pores. By comparing N₂ adsorption and desorption isotherms of MXene@Fe₃O₄ and composite PCMs (Figure S3), we find the pore volume and specific surface area of composite PCMs are both nearly zero, indicating that abundant MA is fully filled into the pores of MXene@Fe₃O₄. The XRD and FTIR spectra of MA, MXene@Fe₃O₄, MXene@Fe₃O₄-MA pristine were shown in Figure S4. MA has characteristic diffraction absorption peaks at 2θ of 22.5° and 24°. The position of XRD characteristic diffraction peak of MXene@Fe₃O₄-MA composite PCMs were basically consistent with that of MA and MXene@Fe₃O₄ (Figure S4B). As seen from FT-IR spectra, the characteristic peaks of -OH at 3415 cm^–1 for MXene@Fe₃O₄, -CH₃ at 2920 cm^–1 and -CH₂ at 2850 cm^–1 for MA were well maintained in MXene@Fe₃O₄-MA, and no new characteristic peaks appeared. These results indicated that only physical interaction occurred without chemical reaction between MA and MXene@Fe₃O₄ (Figure S4A).

Thermal energy storage efficiency is an important evaluating parameter for composite PCMs. The DSC results exhibited that the addition of MXene and Fe₃O₄ can effectively reduce the supercooling degree of MA from 13.78℃ to 10.04℃ (Figure 2A,B). The supercooling is mainly caused by the poor self-nucleation ability of the materials during crystallization, which is not conducive to the storage of latent heat. In this experiment, the in situ chemically synthesized Fe₃O₄ nanoparticles act as nucleating agents in the system, which made MA transform from the original homogeneous nucleation to heterogeneous nucleation, accelerated the crystallization speed, refined the grain structure, and reduced the degree of supercooling.^[35]

Notably, with the increase of Fe₃O₄ content, the phase change enthalpies of MXene@Fe₃O₄-MA composite PCMs loading 70 wt% MA increased from 132.76 to 144.17 J/g (Figure S5), which were increased by 1.04% and 9.60%, respectively, compared with MXene-MA (Figure 2C and Table S1). These results indicated that MXene can be used as a supporting material to encapsulate MA without leakage due to its good adsorption and capillary force. Meanwhile, the addition of Fe₃O₄ can effectively improve the phase change enthalpies of MXene-based composite PCMs. This may be because the adding magnetic Fe₃O₄ nanoparticles can provide heterogeneous nucleation sites compared with the homogeneous nucleation of MA, thus further improving the phase change enthalpy of MXene-based composite PCMs. Therefore, it can be concluded that more magnetic Fe₃O₄ nanoparticles can provide more nucleation sites, thus leading to a higher phase change enthalpy. Furthermore, the phase change temperatures of MXene@Fe₃O₄-MA composite PCMs were almost consistent with the original samples, and the latent heat retention was nearly 100% after undergoing 50 heating/cooling cycles (Figure 2D), indicating the excellent thermal repeatability and stability.

Shape stability and thermal stability are also essential evaluating parameters for the practical application of composite PCMs. Thermal stability of composited PCMs is investigated by TGA (Figure S5). Pristine MA showed totally thermal decomposition from 200 to 300℃. The supporting material (MXene@Fe₃O₄) presented almost no mass loss from 40 to 800℃. The remaining mass fraction of MXene@Fe₃O₄-MA was about 30% when the temperature reached 800℃. Obviously, MXene@Fe₃O₄-MA composite PCMs exhibit high thermal stability below 200℃, which is very important for medium and low temperature practical application. To evaluate the shape stability of pristine MA and MXene@Fe₃O₄-MA composite PCMs, all the samples were placed in an oven at 80℃. All these samples can maintain their original solid shape at room temperature (25℃). When held at 80°C for 25 s (higher than the melting point of MA), pristine MA began to melt and turns into liquid at 40 s, while MXene@Fe₃O₄-MA composite PCMs still maintained the solid shape completely without liquid leakage at 80℃ for 60 min (Figure S6). Additionally, the thermal conductivities of MXene@Fe₃O₄-MA with the ratio of 4:1, 4:2, 4:3 and 4:4 are 0.329, 0.323, 0.321 and 0.315 W/m·K, respectively, which are 49.5%, 46.8%, 45.9% and 43.2% higher than that of pristine MA (0.22 W/m·K) due to the synergistic effect of MXene and Fe₃O₄ (Figure S7). Firstly, MXene is an excellent thermal conductor due to its intrinsic metal-like conductive structure. Secondly, the homogeneously distributed magnetic Fe₃O₄ nanoparticles can realize the radial migration of heat from the surface to the interior, shorten the heat diffusion path, thus further improving the thermal diffusion capacity. Thirdly, the magnetic Fe₃O₄ nanoparticles grown in situ can reduce the interfacial thermal resistance and accelerate the phonon transmission, so as to improve the thermal conductivity (Figure 3F).

Additionally, polarizing optical microscopy (POM) was applied to the in situ observation of crystal evolution of MA, MXene-MA, and MXene@Fe₃O₄-MA composite PCMs during the heating process. As seen from Figure 3C–E, MA exhibited a clear cross-extinction mode spherulite morphology at room temperature. This phenomenon can also be observed in MXene-MA and MXene@Fe₃O₄-MA, but the spherulite size was significantly smaller and the crystal morphology became irregular. This is because MXene and MXene@Fe₃O₄ limited the fluidity of MA molecules in composite PCMs, and the crystallization process of MA molecules in composite PCMs became heterogeneous nucleation from homogeneous nucleation, which refined the grains. However, the presence of MXene and MXene@Fe₃O₄ did not change the crystalline structure of MA molecules in composite PCMs. With increasing temperature, the ordered spherical crystals in MXene-MA and MXene@Fe₃O₄-MA composite PCMs were gradually transformed into disordered amorphous structure until the complete evolution of MA molecules, which were consistent with the evolution of pristine MA, indicating the existence of MXene and Fe₃O₄ nanoparticles did not affect the crystallization behavior of MA molecules.

3.3 Photothermal energy conversion

Inherently, MXene exhibits an excellent electromagnetic wave absorption ability. The absorbed electromagnetic waves can across the MXene lattice structure and undergo internal reflections between the layers, and eventually are absorbed within the structures, which is an essential prerequisite for the photothermal conversion behavior.^[36] The UV-Vis-NIR absorbance spectra of MXene and MXene@Fe₃O₄ were shown in Figure 4B. Compared with pristine MA (Figure S8), MXene and MXene@Fe₃O₄ exhibited almost complete absorption ability in the entire UV-Vis-NIR range. Notably, two enhanced absorption peaks were observed at 680 and 1150 nm due to the LSPR effect of MXene and Fe₃O₄ nanoparticles. Moreover, compared with MXene, the absorption peak of MXene@Fe₃O₄ at 1150 nm was significantly enhanced due to the excellent light absorption of Fe₃O₄ nanoparticles in the near infrared range, which is conducive to the rapid solar harvesting for photothermal conversion and storage of MXene@Fe₃O₄-MA composite PCMs. Benefiting from its metal-like and semi metallic properties, MXene is endowed with significant LSPR effect. Meanwhile, Fe₃O₄ nanoparticles, as photothermal reagents, also have LSPR effect. The synergistic LSPR effect of MXene and Fe₃O₄ nanoparticles can further improve the photothermal conversion performance of MXene@Fe₃O₄-MA composite PCMs (Figure 5A).

Figure 4C presented the photothermal energy conversion curves of MXene@Fe₃O₄-MA composite PCMs under the simulated solar of 0.15 W/cm². When the solar source was switched, the temperature of the prepared composite PCMs rapidly increased from 30℃ to 52℃. When exceeding 52℃, the temperature began to rise slowly between 52℃ to 60℃, corresponding to the photothermal conversion process of MXene@Fe₃O₄-MA composite PCMs, which was consistent with the phase change temperature. The absorbed solar energy was stored in the form of latent heat. Besides, compared with other ratios of MXene@Fe₃O₄ (4:2, 4:3, 4:4), MXene@Fe₃O₄ (4:1)-MA composite PCMs showed a fastest heating rate and reached the phase change point first, indicating that its photothermal conversion rate was the highest. The reason may be that excessive Fe₃O₄ content would agglomerate on the surface of MXene, affecting the photo absorption performance of MXene, thereby weakening the photothermal conversion performance of the composite PCMs. Too little Fe₃O₄ content cannot form a heat transfer channel between the MXene layers, slowed down the thermal diffusion rate, which in turn affects the photothermal conversion performance. Meanwhile, it took only 180 s for MXene@Fe₃O₄ (4:1)-MA composite PCMs to rise temperature from 30℃ to 75℃, which shortened the illumination time by 55% compared with MXene-MA composite PCMs (400 s) under the same light radiation, indicating its potential application value in the field of rapid heating.

As shown in Figure 4D, the photothermal storage efficiencies of MXene@Fe₃O₄ (4:1, 4:2, 4:3, 4:4)-MA composite PCMs were 97.7%, 95.8%, 93.2% and 90.2%, respectively, which were higher than MXene-MA (87.26%). Interestingly, as the content of Fe₃O₄ nanoparticles increases, the photothermal storage efficiency decreases, which was consistent with the results of Figure 4C. This phenomenon can be explained as follows: 1) The synergistic LSPR effect of MXene and Fe₃O₄ nanoparticles enhanced the light absorption ability of MXene@Fe₃O₄-MA composite PCMs; 2) A small amount of Fe₃O₄ nanoparticles in situ anchored in the MXene layers could provide a fast transfer channel for phononic diffusion, and prolong the mean free paths of phonons, thus, improving the heat transfer rate; 3) Excessive Fe₃O₄ nanoparticles tend to agglomerate on the surface of MXene, which conversely reduce the solar-trapping properties of MXene@Fe₃O₄.

To further confirm the synergistic photothermal enhancement effect of MXene and Fe₃O₄ prepared by an in situ synthesis strategy, MXene/Fe₃O₄ composite was prepared by a physical mixing method (Figure S9B). After comparing the photothermal conversion curves, the photothermal conversion efficiency of MXene@Fe₃O₄ (4:1)-MA composite PCMs was better than that of MXene/Fe₃O₄ (4:1)-MA (η = 86.2%). Generally, carbon-based composite PCMs were widely reported for photothermal conversion with a photothermal efficiency from 55.5% to 92.0%.^{[11, 12, 37-40]} Recently, MXene-based composite PCMs gradually were reported for photothermal conversion with a photothermal efficiency from 90.1% to 96.5%.^{[28, 29, 41-43]} Comparatively, our prepared MXene@Fe₃O₄-MA composite PCMs in this study showed a higher photothermal conversion efficiency (97%) under the synergistic effect of MXene and Fe₃O₄ nanoparticles (Figure 4E and Table S2).

Additionally, we also measured the photothermal conversion curves of MXene@Fe₃O₄ (4:3)-MA under different light intensities (Figure S9A). The photothermal conversion ability of composite PCMs were correspondingly improved with the increased light intensity from 0.1 W/cm² to 0.2 W/cm². This is because higher light intensity can induce more photothermal electrons due to LSPR effect, meanwhile, MXene@Fe₃O₄-MA can absorb more light waves per unit time.

Considering the excellent photothermal conversion efficiency of MXene@Fe₃O₄ (4:3)-MA, we mixed it with waterborne elastic polyurethane (PU) to prepare wearable thermal management MXene@Fe₃O₄-MA-PU film. As shown in Figure 4F, the prepared MXene@Fe₃O₄-MA-PU composite film was attached on the skin of the hand, and infrared thermal imaging was used to record the surface temperature evolution of the hand. After being exposed to the sun for about 3 min, the temperature of the composite film increased from 32.8°C to 47.2°C, while the temperature of other parts of the hand had very little change. These results indicated that MXene@Fe₃O₄-MA-PU composite film exhibited excellent photothermal conversion performance in practical applications. In the cold environment, it can be a wearable flexible heating material driven by solar.

To further reveal the synergistic LSPR effect between MXene and Fe₃O₄ nanoparticles and provide a deep understanding of the photothermal transformation phenomenon in MXene@Fe₃O₄-based composite PCMs, the numerical models of Finite Element Method (FEM) based on Maxwell formula were established (Figure 5B–E). In particular, the optical properties of materials are dominated by their LSPR, defined as the collective motions of the conduction electrons induced by light irradiation, which is associated with a considerable enhancement of the electric field. The surrounding electric field intensity of MXene@Fe₃O₄ (4:1) was about 2 times and 8 times higher than that of MXene and pure air when the incident light irradiated. In addition, the electric field intensity of MXene@Fe₃O₄ (4:4) decreased instead, indicating that the agglomeration of magnetic Fe₃O₄ nanoparticles reduced the synergistic LSPR effect between MXene and Fe₃O₄ nanoparticles (Figure 5B–E), which were consistent with the experimental results of photothermal conversion. Therefore, the synergistic LSPR effect between MXene and Fe₃O₄ nanoparticles can be proved by comparing the simulated induced electric field intensity.

3.4 Magnetic-thermal energy conversion

VSM was conducted to evaluate the magnetic property of MXene@Fe₃O₄ (Figure 6B). Obviously, MXene@Fe₃O₄ exhibited an extremely low magnetic retentivity and coercivity, indicating its superparamagnetic nature. These superparamagnetic characteristics mean that an applied magnetic field can attract MXene@Fe₃O₄ while there is no residual magnetism when the magnetic field is removed. Therefore, the magnetic Fe₃O₄ nanoparticles did not tend to aggregate together after removing the magnetic field. Figure 6C showed the temperature-time curves of MXene@Fe₃O₄-MA composite PCMs with different contents of Fe₃O₄ nanoparticles under an alternating magnetic field (the frequency of 1.05 MHz, the alternating current of 10 A). When the alternating magnetic field was applied, the temperature of MXene@Fe₃O₄-MA increased rapidly from room temperature to 50℃ in the form of sensible heat, then the temperature platform appeared on the curves and the generated thermal energy was stored in the form of latent heat. After the phase change process of MA was completed, the temperature further increased rapidly. In addition, the intensity of alternating magnetic field could be enhanced by increasing the intensity of alternating current. It can be seen that the magnetic-thermal conversion capacity increased when the intensity of the alternating magnetic field increased (Figure 6D). These results indicated that MXene@Fe₃O₄-MA exhibited excellent magnetic-thermal conversion and storage capacity under an alternating magnetic field.

The principle of magnetic-thermal effect can be explained as follows: the orientation of the magnetic moment is disordered in the substance without an applied magnetic field according to the Néel and Brownian relaxation effects (Figure 5A). At this time, the magnetic entropy is larger and the adiabatic temperature of the system is lower. After a magnetic field is applied, the magnetic Fe₃O₄ nanoparticles are magnetized, and the molecular magnetic moment tends to be aligned. Resultantly, the magnetic entropy decreases and the adiabatic temperature increases, which in turn releases heat to the environment through heat exchange (Figure 6E).

3.5 Cyclic stability

Cyclic stability of thermal energy conversion and storage is very important for composite PCMs. In this study, our prepared MXene@Fe₃O₄-MA composite PCMs were subjected to 50 photothermal and magnetic-thermal cycles (Figure 7A,B). Resultantly, the photothermal and magnetic-thermal conversion curves after 50 cycles are almost consistent with the original curves, indicating excellent cyclic stability of thermal energy conversion. Additionally, the XRD, FT-IR and DSC results were analyzed after 50 cycles (Figure 7A–F). The phase structures and functional groups types of MXene@Fe₃O₄-MA composite PCMs are almost unchanged. Importantly, the phase change temperature and latent heat of MXene@Fe₃O₄-MA composite PCMs after cycling are almost similar to that before cycling (Figure 7E, F and Table S3). The XRD, FT-IR and DSC results indicate that MXene@Fe₃O₄-MA composite PCMs show superior thermal cycle reliability, which provide a reliable guarantee for pratical application in the heat storage field.