Wide-temperature-range multispectral camouflage enabled by orientation-gradient co-optimized microwave blackbody metastructure coupled with conformal MXene coating

Abstract

Cloaking against electromagnetic detection is a well-researched topic; yet achieving multispectral camouflage over a wide temperature range remains challenging. Herein, an orientation-gradient co-optimized graded Gyroid-shellular (GGS) SiOC-based metastructure with a conformal MXene coating (M@SiOC) is proposed to achieve wide-temperature-range microwave/infrared/visible-light-compatible camouflage. Firstly, the combination of coordinate transformation and genetic algorithm endows the GGS architecture with optimal orientation and gradient, allowing superior microwave blackbody-like behavior. Secondly, a microwave-transparent, low-infrared-emissivity MXene metasurface is constructed in situ to permit wide-temperature-range infrared camouflage. Finally, the outstanding spectral selectivity of MXene enables camouflage against 1.06 μm-lidar and visible-light detection. As a result, the as-fabricated [110]-oriented GGS M@SiOC metamaterials exhibit outstanding wide-temperature-range multispectral camouflage: (i) ultrabroadband microwave absorption exceeding 80% in the X-Ku band from room temperature (RT) to 500 °C with absorption above 86.0% (91.4% on average) at 500 °C; (ii) excellent long-wavelength infrared camouflage for object temperatures from RT to 450 °C, reaching an infrared signal intensity of 78.5% for objects at 450 °C; and (iii) camouflage against both 1.06 μm-lidar and dark environment. Compared with traditional hierarchical metamaterials necessitating complex micro/nano-fabrication processes, this work provides a novel pathway toward the realization of structurally integrated multispectral stealth components by combining flexible metastructure design and high-fidelity additive manufacturing.

---

New concepts

To cope with multicomponent and composite new detection technologies, it is highly desirable to develop a multiband-compatible camouflage technology for both military and civil uses. Widely adopted hierarchical architecture metamaterials are considerably limited due to their high-temperature failure caused by the temperature-induced deterioration of camouflage performances and heterogenous-interface bonding force. This communication provides a feasible structurally integrated wide-temperature-range multispectral-compatible camouflage strategy that combines metastructure design and high-fidelity additive manufacturing with a multifunctional coating. Compared with the traditional complex micro/nano-processing technology, the cost-effectiveness and simplicity of the proposed strategy may promote its widespread use in large-area applications as well as diverse application scenarios, which is of tremendous scientific and tactical relevance.

Introduction

To cope with the rapid development of electromagnetic detectors, research on advanced camouflage is attracting great scientific attention for both military and civil application. Conventional single-band camouflage appears obsolete in view of the pluralistic and multifaceted advancements in modern detection technologies. Current requirements are on the development of multispectral camouflage, i.e., multiband-compatible camouflage technology encompassing microwave (MW) radars, infrared (IR) light, lasers, and visible (Vis) light.^1–3 Yet, due to the mismatch in camouflage principles across the various frequency bands, achievement of multispectral camouflage in the common Vis–MW detection range is an extremely challenging task. Indeed, the camouflage technology should possess the following characteristics: (i) high absorption of Vis light (380–780 nm) and 1.06 μm lasers for camouflage under a dark environment and against near-infrared (NIR) lidars;^4,5 (ii) low IR emittance (high reflection) in the atmospheric window, including the mid-wavelength IR (MWIR, wavelength 3–5 μm) and long-wavelength IR (LWIR, wavelength 8–14 μm) range for camouflage against thermal imaging systems;^2,6 and (iii) high absorption of MWs (e.g., the X-Ku band, 1.67–3.66 cm) for camouflage against MW radar detection.⁷ Fortunately, emerging artificial metamaterials with tailorable electromagnetic properties via flexible subwavelength structure design offer a new approach to achieving multispectral camouflage. A widely adopted strategy is constructing hierarchical multimaterial laminate architectures, e.g., photonic crystals^4,8,9 and stacked metamaterials/metasurfaces,^{1–3,10–12} with feature sizes ranging from nanometers to millimeters. Unfortunately, the application range of such materials is considerably limited due to high-temperature-induced deterioration of interfacial strength, metal oxidation, and temperature-sensitive electromagnetic parameters. Furthermore, the required costly and sophisticated micro/nanofabrication processes, such as printed circuit boards, precision lithography, or high-vacuum thin-film deposition techniques, result in increased difficulties and costs of large-scale fabrication. Hence, structurally integrated temperature-insensitive metamaterials for multispectral camouflage require further scientific investigation.

In our previous work,⁷ graded Gyroid-shellular (GGS) SiOC-based metamaterial microwave blackbodies (MMBs) exhibited superior broadband MW absorption with high-frequency compatibility, temperature insensitivity, and structural scalability, and provided a structural foundation for realizing MW camouflage over a broad temperature range. As aforementioned, IR–MW bi-stealth is not widely achieved due to the difficulty in simultaneously suppressing MW reflection and IR emission. Recent scientific progress focused on the achievement of IR–MW bi-stealth via development of MW-transparent and low-IR-emissivity (MWT/LIRE) frequency-selective surfaces (FSSs), constructed using periodically arranged conductive patches (e.g., metals and indium tin oxide ITO).^{2,3,10,11,13,14} Such MWT/LIRE FSSs can thus be employed to address the aforementioned tradeoff and attain IR camouflage without sacrificing the MW absorption capability. However, conventional MWT/LIRE FSSs with a separated patch array architecture are not suitable for structurally integrated additive manufacturing. According to our previous work,¹⁵ a non-coplanar, separated patch array-like structure can be readily achieved using [110]-oriented Gyroid-shellular (GS) components; this provided the inspiration for in situ construction of MWT/LIRE FSS at the electromagnetic wave (EMW) incident end of GGS-structured MMBs through orientation optimization, thus providing an integrated structure for IR–MW-compatible camouflage. Additionally, the widely reported two-dimensional (2D) MXenes, which exhibit exotic Vis-black-but-IR-white attributes^16,17 as well as great self-assembly characteristics and facile film-forming potential,^18–21 can be processed into a conformal coating on porous or rough substrates to replace traditional metals and ITO towards easy fabrication of MWT/LIRE FSSs based on an orientation-optimized GGS architecture. Furthermore, its unique spectrally selective absorption of NIR lasers and visible light enables camouflage against 1.06 μm lidars and Vis detectors.^2,4

In this work, the combination of an orientation-gradient co-optimized GGS architecture and a conformal MXene coating is harnessed to achieve camouflage covering the Vis-MW spectral range. Initially, based on flexible parametrization design characteristics, the orientation and gradient of the GGS architecture were co-optimized through a coordinate transformation and a genetic algorithm (GA), respectively. Subsequently, high-precision additive manufacturing of polymer-derived SiOC ceramic (PDC-SiOC) enabled high-fidelity fabrication of the orientation-gradient co-optimized GGS architecture, which laid the structural foundation for multispectral-compatible camouflage. Then, based on the optimized GGS architecture, a MXene-based MWT/LIRE FSS was prepared through a facile single-step spray-coating method. A superior wide-temperature-range MW camouflage was achieved due to the unique GGS MMB metastructure featuring high-frequency compatibility, temperature insensitivity, and structural scalability. LWIR camouflage of objects over a wide temperature range was achieved by combination of the directional thermal insulation of the GGS architecture and the low IR emissivity of MXene. Ultimately, the exceptional NIR-Vis spectrally selective absorption of the MXene coating enabled the realization of wide-temperature-range camouflage in both 1.06 μm lidars and dim light or dark environments.

Results and discussion

Orientation-gradient co-optimization of the GGS architecture

A widely adopted strategy for achieving favorable IR low emission and MW-transparent characteristics is through the construction of compact separated patch array structures with a high filling ratio.¹³ The stepwise spiral structure at the EMW incident end of our previously reported [100]-oriented GGS (GGS100) architecture, cannot clearly meet the requirement of a high filling ratio, hence requiring further structural optimization. Fig. 1a shows the orientation optimization procedure undertaken for the GS metamaterial, while accounting for IR–MW-compatible camouflage properties. For the GS with the body-centered cubic (BCC) symmetry, when compared to the typical [100] orientation (i.e., GS100), the [110]-oriented architecture (i.e., GS110) does not exhibit any through holes and features a non-coplanar separated patch array-like structure; features which meet MWT/LIRE FSS structural requirements. Therefore, a conformal MXene-based MWT/LIRE FSS can be easily prepared in situ on GS110 structures using a facile spray-coating approach. Additionally, the parametric design characteristic allows us to directly model the GS110 architecture using a coordinate transformation as follows:¹⁵

|cos(2πx₁₁₀/T) sin(2πy₁₁₀/T) + cos(2πy₁₁₀/T) sin(2πz₁₁₀/T) + cos(2πz₁₁₀/T) sin(2πx₁₁₀/T) | ≤ |c| (1)

where c is the level-set constant; T is the unit cell size of [100]-oriented GS; [x₁₁₀, y₁₁₀, z₁₁₀] equals [x₀, , ], where [x₀, y₀, z₀] denotes the original spatial coordinates. A cuboid unit cell with dimensions of was presented for the GS110 architecture. Here, T was set to 2.54 mm, which is less than one-fifth (∼3.32 mm) of the shortest wavelength of the X-Ku band (16.7 mm at 18 GHz), hence meeting the subwavelength-size criteria.²²

Fig. 1 (a) Orientation optimization of the BCC GS and corresponding structural configurations viewed from different crystal orientations;¹⁵ (b) GA-enabled gradient optimization of the GGS110 architecture, with the discretized GGS110-2.0 architecture constructed using cubic voxels with different edge sizes (left), and views from different directions (right); for the GGS110 SiOC MMBs at different heights, (c) globally optimal |c|–z curves and (d) corresponding SAF–z curves; and (e) simulated MW response spectra of the SiOC MMBs with the GGS110-2.0-z_opt and GGS110-2.0-z^1/3 architectures. The inset in (e) depicts the correspondence diagram of the final optimized GGS110-2.0-z_opt (yellow) and GGS110-2.0-z^1/3 (blue) geometry configurations.

As we reported previously,⁷ varying the c curvature along the EMW propagation direction (the z-axis here), permits the tailoring of the impedance matching of the MMBs, thus optimizing the electromagnetic response characteristics in the MW range. Through a trial and error process, it is concluded that an architecture with a z^1/3 curvature exhibits the optimal MW absorption performance. Herein, based on the same SiOC substrate (the relative permittivity as a function of frequency is shown in Fig. S1a, ESI†), a GA was adopted to obtain the globally optimal c curvature of the GGS110 architecture; the optimization was performed so as to simultaneously maximize the absorption and minimize the reflection over the entire X-Ku band. As illustrated in the left-hand side of Fig. 1b, the complex-shaped GGS110 architecture, with a unit cell size of (n = 2.0, 2.5, and 3.0), can be discretized as an equivalent architecture assembled using subwavelength-size cubic voxels to considerably simplify the modeling process in the electromagnetic simulation. The smooth GGS110 structure can be achieved by gradually decreasing the voxel edge size (from T/10 to T/40 here), and the globally optimal c curvature can be obtained based on the discretized GGS110 architecture (see Supporting Note 1 for the detailed optimization process, ESI†). The right-hand side of Fig. 1b presents the as-obtained optimal GGS110 architecture with a height of (GGS110-2.0-z_opt) and related views from different directions. The top view at the EMW incident end clearly shows a non-coplanar separated patch array-like structure, while the other views reveal graded through-hole features. Compared with conventional closed-cell shell-based microlattices, the interconnecting open cells of the GGS110 architecture make the interior space accessible for the subsequent spray-coating of MXene. To be consistent with our previous work,⁷ |c| was set to be within the range of 0.05–1.067; the final optimized c variation curves for the GGS110 structure at different heights of (n = 2.0, 2.5, and 3.0, which correspond to the GGS110-2.0, GGS110-2.5, and GGS110-3.0 architectures, respectively) are given in Fig. 1c, with the corresponding model schematic in the inset and fitted quadratic equations as follows:

c(z)_n = 2.0 = 0.0844 (z/T)² – 0.3596(z/T) + 0.3897, (2)

c(z)_n = 2.5 = 0.0812(z/T)² – 0.2877(z/T) + 0.3049, (3)

c(z)_n = 3.0 = 0.0565(z/T)² – 0.2397(z/T) + 0.3043. (4)

For comparison, the analytical equation of the c(z^1/3) curvature is also provided:⁷

Fig. 1d depicts the solid area fraction (SAF) of the cross sections perpendicular to the z-axis for all GGS architectures, within the range of 3.427%–78.763%; these curves exhibit the same trend with the c curves and demonstrate the efficacy of tuning the impedance gradient by variation of the c curvature.⁷ In view of the lightweight requirements, the GGS110-2.0 architecture was chosen, and the effective thickness (d_eff, which is equal to that of bulk structures with an identical solid volume) was calculated as 2.09 and 2.63 mm for the GGS110-2.0-z_opt and GGS110-2.0-z^1/3 architectures, respectively, indicating that GGS110-2.0-z_opt exhibits lightweight characteristics with an ultralow surface density of 3.55 kg m⁻², superior to 4.47 kg m⁻² of GGS110-2.0-z^1/3. From the simulated MW response spectra (Fig. 1e and Fig. S1b, ESI†), namely the reflectivity (ρ_MW), absorptivity (α_MW), and transmissivity (τ_MW) as a function of frequency, it can be observed that the blackbody-like MW absorption behavior of GGS110-2.0-z_opt is less sensitive to frequency than that of GGS110-2.0-z^1/3, with an average ρ_MW/α_MW ([small rho, Greek, macron]_MW/[small alpha, Greek, macron]_MW) of 5.0%/87.8% and a lower low-frequency reflectance, which is significant given that the conductive MXene coating can further improve the reflectance. The electromagnetic response is closely related to the SAF distribution characteristics. As shown in Fig. 1d, the GGS110-2.0-z_opt architecture exhibits a larger SAF than the GGS110-z^1/3 architecture at the upper end (z > 0) and a smaller SAF at the lower end (z < 0), which can be also evidenced by the model comparison shown in the inset of Fig. 1e. To summarize, with better MW absorption capacity, the orientation-gradient co-optimized GGS110-2.0-z_opt architecture is preferred over the previously reported GGS110-2.0-z^1/3 architecture and lays the structural foundation for IR-MW-compatible camouflage.

Material characterization of GGS110-2.0-z_opt M@SiOC

Based on the obtained orientation-gradient co-optimized GGS110-2.0-z_opt architecture, a facile spray-coating approach can be utilized to deposit an MXene coating on the porous GGS SiOC substrate for achieving multispectral camouflage. The fabrication process of GGS-2.0-z_opt SiOC with conformally coated MXene (M@SiOC) is illustrated in Scheme 1, while the conformal MXene coating itself was fabricated by spraying an aqueous solution of Ti₃C₂T_x MXene onto the hydrophilic GGS110-2.0-z_opt SiOC substrate using an air spray gun.

Scheme 1 Schematic of the fabrication process of GGS110-2.0-z_opt M@SiOC using the spray-coating method.

Our previously reported high-precision PDC-SiOC slurry, which can be used to realize centimeter-size additive manufacturing of ceramic-based metamaterials while retaining of ultrahigh printing resolution of 20 μm, enables the high-fidelity fabrication of the as-optimized GGS110-2.0-z_opt architecture.^7,23Fig. 2a shows photographs of the GGS110-2.0-z_opt samples for electromagnetic testing in the X-Ku band, at different stages including the orange-colored precursor structure (before pyrolysis), black-colored SiOC pyrolyzed at 1300 °C (SiOC-1300), and dark-gray-colored M@SiOC, with a uniform linear shrinkage of approximately 41.2% for the SiOC samples after the pyrolysis process. Additionally, the color of the M@SiOC samples turned to a uniform dark gray, indicating a homogeneity of the spray-coated MXene (SC-MXene) coating. As shown in Fig. 2b, the static contact angle between the SiOC substrate and the aqueous MXene dispersion was determined to be 46.7° ± 0.1°; such a high hydrophilicity facilitates the realization of a uniform MXene coating on the SiOC substrate. For the SiOC precursor samples, the pyrolysis temperature was set to 1300 °C to achieve a compromise between good mechanical properties and high dielectric loss. According to our earlier findings,^7,15 the phase components of SiOC-1300 primarily consist of the β-SiC phase, cristobalite (c-SiO₂) phase, and conductive turbostratic carbon (t-C), all of which are homogeneously distributed in the amorphous SiOC matrix.

Fig. 2 (a) Optical photographs of the GGS110-2.0-z_opt samples at different stages for electromagnetic testing in the X-Ku band; (b) optical tensiometry image of aqueous MXene dispersion on the SiOC substrate (top), and digital images of the bulk SiOC samples with and without the MXene coating (bottom); (c) XRD patterns of MXene with the SEM image in the top inset; (d) TEM and (e) HRTEM images of the MXene nanosheets with the SAED pattern in the inset; (f) the side view and (g) the top view of GGS110-2.0-z_opt M@SiOC at different magnifications; (h) corresponding SEM–EDS mapping of (g3); (i) cross-sectional SEM image of the SC-MXene coating and corresponding SEM–EDS mapping; and (j) and (k) morphology of the evenly distributed MXene coating and the zoomed-in fracture surface, respectively.

The X-ray diffraction (XRD) patterns in Fig. 2c show a pronounced shift of the (002) peak from the original angle of 9.7° of the hexagonal Ti₃AlC₂ (PDF#52-0875) to the angle of 6.4° of the delaminated Ti₃C₂T_x; this shift, alongside the distinct 2D lamellar microstructure depicted in the inset, indicates the successful synthesis of layered Ti₃C₂T_x MXene.^17,24–26 Furthermore, the X-ray photoelectron spectroscopy (XPS) spectra of Ti₃C₂T_x MXene in Fig. S2 (ESI†) confirm that the Ti : C : O : F atomic ratio is 37.5 : 23.0 : 22.9 : 16.6, and its functional groups contain a mixture of oxide (–O–), hydroxyl (–OH) and fluorine (–F) terminations with the simple structure Ti₃C₂O_0.5 (OH)_1.5 F_1.5. The transmission microscopy (TEM) images in Fig. 2d show the micromorphology of the few-layer Ti₃C₂T_x MXene nanosheets, which have a lateral size of 1.4 μm. Furthermore, the monolayer thickness can be determined as ∼1.68 nm by atomic force microscopy (AFM) imaging as per Fig. S3 (ESI†). Lattice fringes with an interplanar spacing of 0.23 nm visible in the high-resolution TEM (HRTEM) image (Fig. 2e), correspond to the (103) plane of Ti₃C₂T_x MXene,^27–30 while the associated selected-area electron diffraction (SAED) patterns reveal a typical hexagonal structure.^26,31

Fig. 2f (corresponding to view #2 in Fig. 1b) and Fig. S4 (ESI†) show the side views of the GGS110-2.0-z_opt M@SiOC samples at different length scales, revealing a longitudinal-gradient structure with transverse periods of 3.59 and 2.54 mm and a height of 7.18 mm, respectively. The morphology at the EMW incident end (Fig. 2g1, corresponding to view #1 in Fig. 1b) clearly reveals the non-coplanar separated patch array-like architecture with a repeating period of 3.59 mm × 2.54 mm, which functions as the MWT/LIRE FSS. As shown in Fig. 2f3, g2 and Fig. S4c (ESI†), the topmost shell thickness is as low as approximately 20 μm, indicating the high precision and high shape fidelity of the additively manufactured GGS110-2.0-z_opt SiOC, which guarantees the reliability of predicting the MW response through simulations. In Fig. 2g3, i, and j, it can be observed that the uniformly distributed SC-MXene coating with an ultrathin thickness of ∼2 μm is firmly attached to the SiOC substrate due to the self-assembly deposition behavior. The magnified view of the fracture surface in Fig. 2k shows that the SC-MXene coating is composed of stacked delaminated Ti₃C₂T_x nanosheets and is fluffier than the vacuum-filtered MXene (VF-MXene) films (Fig. S5, ESI†).

Wide-temperature-range MW absorption performance

As shown in Fig. 3a, a transmission/reflection method using rectangular waveguides was employed herein to evaluate the wide-temperature-range MW response of the materials in the X-Ku band (8.2–18.0 GHz).^7,32,33Fig. 3b depicts clear non-resonant blackbody-like absorption spectra for GGS110-2.0-z_opt M@SiOC and the reference SiOC at RT. Therein, an increasing trend of absorption with frequency is noted, with respective [small rho, Greek, macron]_MW/[small alpha, Greek, macron]_MW values of 12.2%/87.6% and 5.5%/91.6%. Due to the deteriorating impedance matching caused by the conductive MXene coating, GGS110-2.0-z_opt M@SiOC exhibited a slightly higher reflection and a weaker absorption than the reference SiOC, but α_MW still exceeds 80% over the entire X-Ku band. For convenience, the total electromagnetic interface (EMI) shielding efficiency (SEt) can be expressed as −10 log 10 (τ_MW). By examining Fig. 3c it is observed that the MXene coating endows GGS110-2.0-z_opt M@SiOC with a greater EMI shielding capacity than the reference SiOC, with SEt ≥ 24.0 dB (27.3 dB on average) over the entire X-Ku band, exceeding the value of 20 dB required for commercial applications.²⁶ Furthermore, for the GGS110-2.0-z_opt SiOC sample, the good agreement observed between the experimental data and the simulation results (Fig. 1e) further proved the high shape fidelity of the additively manufactured SiOC metastructure.

Fig. 3 (a) Schematic of the setup used for testing the MW response in a broad temperature range; experimentally determined frequency dependences for the GGS110-2.0-z_opt SiOC and M@SiOC samples at RT of: (b) ρ_MW and α_MW, (c) τ_MW (SEt); wide-temperature-range experimental values of GGS110-2.0-z_opt M@SiOC: (d) α_MW & ρ_MW and (e) τ_MW (SEt) during the heating process from 100 °C to 500 °C; (f) α_MW & ρ_MW and (g) τ_MW (SEt) during the cooling process from 500 °C to 100 °C; (h) summary of the [small rho, Greek, macron]_MW/[small alpha, Greek, macron]_MW/SEt_ave values versus temperature in a heating–cooling cycle; and (i) schematic of the MW absorption mechanism for GGS110-2.0-z_opt M@SiOC.

Further research was conducted on the MW response of GGS110-2.0-z_opt M@SiOC over a wide temperature range during heating (100 °C–500 °C, Fig. 3d and e) and cooling (500 °C–100 °C, Fig. 3f and g). Non-resonant blackbody-like absorption behavior was observed here as well, with α_MW ≥ 80% over the entire X-Ku band; in particular, α_MW still exceeds 86.0% (91.4% on average) at 500 °C. As demonstrated in Fig. 3h, as the temperature varies during the heating–cooling process, the SEt and reflectivity spectra of GGS110-2.0-z_opt M@SiOC exhibit a gradually decreasing trend, while the absorption spectrum shows an upward trend. All these spectra exhibit good temperature-insensitive MW absorption performance with [small rho, Greek, macron]_MW < 12.7%, [small alpha, Greek, macron]_MW > 86.8%, and SEt_ave > 16.9 dB over the range from RT to 500 °C. For the GGS110-2.0-z_opt M@SiOC sample, after testing at 500 °C, the peeled-off coating from the SiOC substrate is visible in the SEM image in Fig. S6 (ESI†), with the coating structure changing from a lamellar stacked architecture (Fig. 3k) to a dense and compact morphology. An MXene-to-TiO₂ phase change is evident from the XRD patterns of the SC-MXene films preoxidized at 500 °C in Fig. S7 (ESI†), where the representative (002) peak of Ti₃C₂T_x almost vanishes, while a prominent peak of the TiO₂ phase (including rutile and anatase) appears.

To clarify the influence of the MXene coating on the MW absorption performance, the sheet resistance (R_s) of the MXene coating after pre-oxidation at different temperatures was investigated and results are shown in Fig. S8 (ESI†). At around 250 °C–300 °C, a noticeable transition from conductive to insulating behavior occurs, implying a major alteration in the inherent MW absorption mechanism. Two competing mechanisms exist during the heating process: one is MXene oxidation, which causes a decrease in conductivity and can therefore improve the impedance matching at the EMW incident end; the other is the elevated high-temperature relaxation of the SiOC substrate, which deteriorates the surface impedance matching and results in a better EMI shielding performance.^7,15 The MXene coating dominates the MW absorption capability below 250 °C due to the strong ohmic loss, while the SiOC substrate becomes increasingly important above 300 °C. Even at 500 °C, the MXene-to-TiO₂ transformation improves the impedance matching characteristics, and the high-temperature relaxation of the SiOC substrate can induce a better EMW attenuation capability; therefore, the overall enhanced MW absorption performance is retained for GGS110-2.0-z_opt M@SiOC, and it is superior to that of M@SiOC and the reference SiOC at RT. Additionally, the material composition remains constant during the cooling process, and the MW absorption performance steadily improves as the temperature decreases, exhibiting a trend similar to that of the reported SiOC MMBs.⁷

For the SC-MXene coating, R_s was measured as 0.8836 kΩ sq⁻¹ at RT (Fig. S8, ESI†), and the conductivity (σ) was calculated to be ∼565.8 S m⁻¹ for the 2-μm-thick MXene coating according to the relationship σ = 1/(R_st), where t denotes the thickness. According to the relationship ,^20,34 where f denotes EMW frequency, and μ is equal to the permeability of vacuum (μ₀), the skin depth (δ) at MW frequencies of 8.2–18.0 GHz was calculated to be 158.1–234.0 μm, which is significantly greater than the material thickness of 2 μm. Therefore, the SC-MXene coating with a sub-skin-depth thickness can be easily penetrated by the MWs. The schematic of the MW absorption mechanism for GGS110-2.0-z_opt M@SiOC is shown in Fig. 3i. When working at MW frequencies, GGS110-2.0-z_opt M@SiOC can be regarded as the frustum of a cone array configuration comprised of the SiOC substrate coated with SC-MXene with a sub-skin-depth thickness, and its exceptional MW absorption performance stems from the good impedance matching and high MW attenuation capability. As demonstrated in our previous work,⁷ the MMBs resemble an MW trap in which MWs can easily enter but their exit is hindered; the stepwise GGS metastructure contributes significantly more to the total impedance matching than the constituent materials do. The SC-MXene coating promotes the attenuation of the EMWs primarily owing to its strong ohmic loss and the multiple reflections occurring at the MXene/SiOC and MXene/air heterophase interfaces, while the GGS architecture improves the impedance matching of the materials to further enhance the MW absorption capability. By harnessing the high-temperature relaxation and multiscale polarization losses of the dielectric SiOC substrate,²³ a remarkable wide-temperature-range camouflage against MW radars is eventually demonstrated.

Wide-temperature-range IR-Vis camouflage performance

LWIR Camouflage Performance. The LWIR camouflage performance of objects over a wide temperature range is evaluated in terms of the radiation temperature (T_r) reduction.²Fig. 4a illustrates the schematic setup used for wide-temperature-range IR testing, where objects at various temperatures are mimicked by a heating stage and an LWIR camera was used to record the radiation temperature; the samples were placed on the heating stage and were comprised of 5 × 7 unit cells with dimensions of 18.0 (length) × 17.8 (width) × 7.2 (height) mm³. The LWIR images of GGS110-2.0-z_opt M@SiOC at different heating temperatures (equivalent to the object's temperature, i.e., T_obj) are shown in the upper right corner of Fig. 4b, exhibiting high-performance LWIR camouflage, while their corresponding radiation temperature is shown in Fig. 4c. For objects at temperatures of 100 °C/450 °C (373.15/723.15 K), the average radiation temperature (T_{r, ave}) of GGS110-2.0-z_opt M@SiOC was recorded as 48.4 °C/219.5 °C (321.55/492.65 K), which corresponds to a large reduction of 51.6%/51.2%, respectively. Furthermore, SEM micrographs in Fig. S9 (ESI†) demonstrate that the MXene coating at the EMW incident end is firmly adhered onto the SiOC substrate without signs of peeling off and maintains a lamellar structure after testing at 450 °C, a finding which indicates an excellent thermal stability. According to the Stefan–Boltzmann law, namely P = εσT(K)⁴ (where σ is the Stefan–Boltzmann constant), the thermal radiation intensity (P) is proportional to the surface emissivity (ε) and the fourth power of the surface temperature (T(K)).^35,36 Therefore, for objects at 100 °C/450 °C, the IR signal intensity can be enormously reduced by 44.9%/78.5% using GGS110-2.0-z_opt M@SiOC, which is related to the structural thermal insulation (ΔT) and IR emissivity of the MXene coating.

Fig. 4 (a) Schematic of the setup used for the IR camouflage test; for GGS110-2.0-z_opt M@SiOC and the reference BP@SiOC: (b) LWIR camera images (right) of the samples held at different temperatures and corresponding digital images (left), and (c) measured LWIR radiation temperature, with the inset depicting the lateral LWIR image of GGS110-2.0-z_opt BP@SiOC held at 450 °C; (d) absorptance/emissivity spectra of SiOC pyrolyzed at 900 °C–1300 °C, SC-MXene at RT or after 250 °C pre-oxidation, and reference VF-MXene at RT; (e) XRD patterns of SC-MXene after pre-oxidation at different temperatures; (f) optical (left) and LWIR (right) images of SC-MXene held at 50 °C; (g) experimentally determined absorptance/emissivity spectra over 0.38–1.2 μm wavelength for SiOC-1300 as well as SC-MXene at RT or after pre-oxidation at 250 °C; and (h) first-principles calculations of the reflectivity spectra of Ti₃C₂T_x with various terminations and interlayer spacings over 0.38–7.0 μm wavelength.

A reduction in the surface temperature is desired for realizing IR camouflage.^6,37 However, due to the micron-sized interwoven pores of GGS110-2.0-z_opt M@SiOC (Fig. 2g), it is difficult to directly monitor the surface temperature using thermocouples. To quantify the contribution of the thermal insulation to the IR camouflage, GGS110-2.0-z_opt SiOC spray-coated with a commercial high-temperature-resistant black paint (BOTNY B-1200, with a constant ε of 0.97),^38,39 here denoted GGS110-2.0-z_opt BP@SiOC, was utilized as a reference sample to derive the actual temperature from the LWIR images in Fig. 4b. Notably, a clear dark dot array is visible in the high-temperature LWIR images of GGS110-2.0-z_opt BP@SiOC, which corresponds to the topmost low-temperature region and originates from the non-coplanar separated patch array-like architecture of GGS110 at the EMW incident end. In Fig. 4c, at the highest T_obj of 450 °C, the lowest surface radiation temperature (T_{r, min}, corresponding to the dark spots in Fig. 4b) is as low as 252.5 °C (the average surface radiation temperature, T_{r, ave}, is 299.7 °C) for GGS110-2.0-z_opt BP@SiOC. Furthermore, the plausible thermal insulation performance can also be highlighted by the lateral LWIR image in the inset of Fig. 4c. For GGS110-2.0-z_opt M@SiOC, the low surface temperature significantly contributes to the stable IR emission of the topmost MXene coating, which plays a major role in IR camouflage. The existing literature demonstrates that GS with cubic symmetry has an isotropic effective thermal conductivity, which varies linearly with the relative density (equivalent to the SAF in Fig. 1d).^40,41 Thus, the good directional thermal insulation of GGS110-2.0-z_opt M@SiOC mainly originates from the non-through-hole longitudinal gradient architecture of GGS110 and the low intrinsic thermal conductivity of SiOC.⁴²

The IR camouflage performance is also closely related to the IR emission characteristics of the material. As observed in Fig. 4b, due to the presence of the conformal MXene coating, the LWIR image of GGS110-2.0-z_opt M@SiOC appears cooler than that of GGS110-2.0-z_opt BP@SiOC. Kirchhoff's law states that infrared ε (ε_IR) is equal to absorptance (α_IR), which can be calculated as ε_IR = 1 – Reflectivity for IR-opaque MXene/SiOC materials.⁴³ The experimentally determined α_IR/ε_IR spectra in Fig. 4e demonstrate that SiOC has a high IR emissivity while MXene has a low IR emissivity (i.e., IR white);^16,17 therefore, the IR camouflage performance of GGS110-2.0-z_opt SiOC can be substantially improved by utilizing the conformal MXene coating. The average MWIR and LWIR ε values at RT are 33.0% and 23.1% for SC-MXene, respectively, which are greater than that of VF-MXene (19.4% and 10.7%) due to the larger interlayer spacing of the former (Fig. 2k and Fig. S5, ESI†).¹⁷ SiOC pyrolyzed at various temperatures (namely 900 °C, 1100 °C, and 1300 °C; here denoted SiOC-900, SiOC-1100, and SiOC-1300, respectively) exhibits different IR emission characteristics, with SiOC-1300 dominating the best LWIR emissivity (76.4% on average). Additionally, a deep α_IR/ε_IR valley occurs at around 9.0 μm for SiOC; it is speculated that such a valley is related to the stretching of the Si–O–Si bonds, which is attributed to the absorption peak at about 1100 cm⁻¹ (9.09 μm) in the Fourier-transform infrared spectroscopy (FT-IR) spectra in Fig. S10 (ESI†).^44,45

Investigation of the relationship between the thermal stability and the IR emission performance is vital because MXene is known before its thermal instability and susceptibility to oxidation.^16,46Fig. 4d shows that the MWIR/LWIR ε are 83.8% and 49.5% for SC-MXene after pre-oxidation at 250 °C, respectively, a finding which indicated that the MWIR camouflage ability deteriorates significantly at elevated temperatures while the LWIR camouflage characteristics are retained. Fig. S11 (ESI†) presents the LWIR images and corresponding T_r values for the bulk M@SiOC, SiOC, and reference BP@SiOC at different heating temperatures. The LWIR image of the bulk SiOC is similar to that of the reference BP@SiOC in the range from RT to 300 °C, while the bulk M@SiOC exhibits excellent wide-temperature-range LWIR camouflage performance. For objects at 300 °C, the radiation temperatures are 146.7 °C and 180.5 °C for GGS110-2.0-z_opt M@SiOC and the bulk M@SiOC, respectively, and the difference between the two structures lies in the high thermal insulation of the GGS110-2.0-z_opt architecture. In addition, both SiOC and SC-MXene after pre-oxidation at 250 °C possess high emissivity values (91.2% and 75.7%, respectively) in the non-atmospheric window (wavelengths 5–8 μm); thus, the heat radiating from high-temperature objects can also be well dissipated via radiative cooling due to the large specific surface area of the GGS architecture, contributing to good thermal stability as well.²

The phase information of SC-MXene after pre-oxidation at various temperatures was further characterized via XRD, and the corresponding XRD patterns are shown in Fig. 4e, where an MXene-to-anatase-to-rutile transformation can be observed. For SC-MXene after pre-oxidation at 300 °C, the (002) characteristic peak almost vanishes with the emerging anatase phase, a finding which compares favorably with the aforementioned conductive-to-insulating transition behavior occurring at around 250–300 °C (Fig. S8, ESI†). Interestingly, the structural thermal insulation ensures that the surface temperature of GGS110-2.0-z_opt M@SiOC is lower than 250 °C even at the highest T_obj of 450 °C (Fig. 4c), hence resulting in a good LWIR camouflage performance for high-temperature objects owing to the residual MXene phase. Notably, the magnified views of the XRD patterns reveal that the (002) characteristic peak gradually shifts toward higher 2θ angles with increasing pre-oxidation temperature. Based on Bragg's equation 2d sin θ = nλ, it can be speculated that the interplanar spacing d decreases steadily because of the MXene-to-TiO₂ transformation. Fig. 4f shows the LWIR images at T_obj = 50 °C (right) and the corresponding digital images (left) of MXene sprayed on a glass substrate after pre-oxidation at various temperatures; a piece of electrical tape (3M Scotch Super 88 Vinyl, with a known emissivity of 0.95⁴⁷) was used as the reference blackbody. It is evident that the LWIR camouflage ability gradually degrades with increasing pre-oxidation temperature but remains still good after pre-oxidation at 250 °C.

Vis-1.06 μm-laser camouflage performance. The Vis-NIR absorption spectra of SiOC-1300 as well as SC-MXene at RT or after pre-oxidation at 250 °C were acquired to confirm their camouflage performance against Vis light and 1.06-μm laser and are shown in Fig. 4g. In the Vis band, SiOC-1300 has an absorptivity (α_Vis) of 95.7%, while SC-MXene has an α_Vis of 92.7% at RT and of 91.2% after pre-oxidation at 250 °C. Therefore, both materials exhibit a distinct black-colored appearance due to their high Vis-light absorption characteristics, which can be further confirmed by the corresponding optical images in Fig. 2a (middle), 4b (left upper) and 4f (left). According to optical images in Fig. 4f, the SC-MXene after pre-oxidation at temperatures from 50 °C to 250 °C shows a distinct black-colored appearance and adheres firmly to the glass substrate with no sign of peeling. As a result, the combination of the intact black-colored SC-MXene coating and the non-through-hole GGS110 SiOC substrate provides the target beneath the GGS110-2.0-z_opt M@SiOC samples with a black-colored camouflage appearance, as shown in Fig. 2a (middle) and 4b (left upper). Thus, the target outline can be well hidden under dim light or dark environments, thereby effectively reducing the probability of detection and recognition by the human eye or photo-reconnaissance satellites.^2,4,17 For active 1.06 μm lidar detection, low NIR reflection (high absorption) characteristics near 1.06 μm are required to reduce the echo power of incident lasers. SiOC-1300 demonstrates a highly efficient (98.4%) 1.06 μm laser absorption, and for SC-MXene the efficiency is 94.5%/92.5% at RT or after pre-oxidation at 250 °C, respectively. Therefore, due to its good structural thermal insulation, GGS110-2.0-z_opt M@SiOC can be expected to achieve satisfactory wide-temperature-range Vis-1.06 μm-laser-compatible camouflage. Furthermore, the stepwise gradient structure with a curved surface appearance can reduce the Vis-NIR reflection, thus improving the Vis-1.06 μm-laser camouflage performance even further.

First-principles calculations of the Vis-IR reflectivity. As shown in Fig. 4h, for Ti₃C₂T_x MXene with different terminal groups and different interlayer spacings (d), first-principles calculations were employed to further study the Vis-IR reflectance characteristics over the wavelength 0.38–7.0 μm. It was found that the Vis-IR electromagnetic response is significantly influenced by both the functional group type and the interlayer spacing. At a given interlayer spacing of d = 20 Å, fluorine-terminated (–F) Ti₃C₂T_x exhibits the best spectral selectivity, followed by –OH-terminated Ti₃C₂T_x, which shows a high Vis-1.06 μm-laser absorption (low reflection) and a low MWIR emission (high reflection), while –O-terminated Ti₃C₂T_x exhibits the worst performance. It can be concluded that oxidation significantly worsens the spectral selectivity of Ti₃C₂T_x MXene, consistent with the abovementioned experimental data (Fig. 4d and g). Our prepared Ti₃C₂T_x with an O : OH : F blend ratio of 1 : 3 : 3 has a spectral selectivity that is second only to that of –F-terminated Ti₃C₂T_x but superior to that of –OH-terminated Ti₃C₂T_x. For Ti₃C₂T_x with different interlayer spacings (d = 20, 50, and 100 Å), it can be observed that the overall reflectivity gradually decreases with the widening of the interlayer spacing, enhancing the Vis-NIR absorption but deteriorating the IR camouflage performance, in line with the above experimental findings (Fig. 4d). Therefore, the fluffy SC-MXene possesses an elevated absorption for Vis/1.06 μm laser at the expense of a slightly higher IR emission compared with the more compact VF-MXene. Furthermore, an abrupt absorption-to-reflection transition, which corresponds to the dielectric-to-metallic transition, is observed at around 1.3 μm wavelength, which is the reason why Ti₃C₂T_x MXene has good spectral selectivity.¹⁷

In conclusion, the remarkable Vis/1.06 μm/LWIR-compatible camouflage performance of GGS110-2.0-z_opt M@SiOC for object temperatures in the range from RT to 450 °C was experimentally demonstrated and theoretically analyzed. Such a favorable performance stems from the combination of the excellent thermal insulation and superior spectral selectivity of MXene, together with the radiative cooling offered by the SiOC substrate and the oxidized MXene in the non-atmospheric window.

Wide-temperature-range multispectral camouflage mechanisms

As shown in Table S1 (ESI†), the GGS110-2.0-z_opt M@SiOC metamaterials with an integrated structure exhibits the best all-around performance compared with the reported hierarchical architecture metamaterials. The schematic in Fig. 5 elucidates the underlying working mechanisms of the remarkable wide-temperature-range multispectral camouflage. As illustrated in the upper inset, GGS110-2.0-z_opt M@SiOC can be considered equivalent to a MXene-based MWT/LIRE FSS integrated with a GGS MMB metastructure, with a different working mechanism for EMWs of different frequencies is different. As shown in the lower-left corner of the figure, at MW frequencies, the non-coplanar separated patch array-like surface of the GGS110 structure with conformally coated MXene is similar to metallic gratings with narrow slits, which are highly transparent to MWs, hence providing good impedance matching. As shown previously in Fig. 3i, the GGS110-2.0-z_opt MMB architecture with high-frequency compatibility, temperature insensitivity, and structural scalability enables the realization of a remarkable wide-temperature-range ultrabroadband MW absorption. Unlike the traditional metal-backed resonant MW absorbers,^48,49 which are susceptible to variations in the parent body composition and heavily rely on destructive interference, the non-resonant MW blackbody-like absorption characteristics are hardly affected by parent body composition and back-panel material type (either non-metal, blank, or metal).⁵⁰

Fig. 5 Schematic overview of the multispectral camouflage mechanisms of GGS110-2.0-z_opt M@SiOC over a wide temperature range.

As shown in the lower central part of Fig. 5, for MWIR/LWIR camouflage, the topmost MXene coating serves as a low-emission surface, while the GGS110 architecture attached to the high-temperature object serves as a thermal insulator. Both the unique non-through-hole GGS110 structure and the IR-opaque SiOC substrate meet the IR-opacity requirement of thermal insulators. When combined with the intrinsically low MWIR/LWIR emission of the topmost MXene coating, the thermal radiation of high-temperature objects is considerably reduced by GGS110-2.0-z_opt M@SiOC, hence enabling the achievement of IR camouflage over a broad temperature range. Additionally, the radiative cooling of the SiOC substrate and the MXene oxidation in the non-atmospheric window endow the porous GGS110 structure with better heat dissipation characteristics than the traditional pore-free, dense metamaterials of laminated architecture, thus ensuring a good thermal stability during prolonged exposure to high-temperature environments.

For conventional IR camouflage coatings primarily composed of metallic microparticles, both the high glossiness and high brightness limit their compatibility with Vis light and NIR lasers.⁵¹ Here, the good Vis-NIR absorption characteristics of SiOC and MXene endow the prepared structures with correspondingly good Vis-NIR absorption characteristics. Furthermore, the micron-scale curved surface and the high surface area of the porous GGS structure promote the occurrence of more scattering events and reflections of the incident Vis-NIR light, which further enhance the Vis-1.06 μm-laser camouflage performance. Additionally, the combination of the good thermal insulation of the GGS110 structure and the good thermal stability of SiOC and SC-MXene enables excellent wide-temperature-range camouflage performance under a dark environment and against 1.06 μm lidars.

Conclusions

Based on our previously reported GGS100 architecture,⁷ an integrated structure based on the coupling of an MMB metastructure with an MWT/LIRE FSS was realized for the first time through the synergy of an orientation-gradient co-optimized GGS110 macrostructure and conformal MXene coatings with sub-skin-depth thickness. Ultimately, the high-fidelity additively manufactured GGS110-2.0-z_opt M@SiOC metamaterials exhibited remarkable wide-temperature-range multispectral-compatible camouflage performance: (i) blackbody-like ultrabroadband MW absorption exceeding 80% in the X-Ku band from RT to 500 °C; in particular, the absorption exceeds 86.0% (91.4% on average) at 500 °C; (ii) excellent LWIR camouflage for objects with temperatures ranging from RT to 450 °C, reaching an IR signal intensity reduction of 78.5% (which corresponds to a temperature reduction of 230.5 °C) for objects at 450 °C; and (iii) camouflage against both 1.06 μm wavelength lidars and dim light or dark environments. Our study provides a feasible and novel multispectral-compatible camouflage strategy that combines integrated metastructure design and high-fidelity additive manufacturing with a multifunctional coating, which is of tremendous scientific and tactical relevance. Compared with the conventional complex micro/nanoprocessing technology, the cost effectiveness and simplicity of the proposed strategy may promote its widespread use in large-area applications as well as diverse application scenarios.

Furthermore, it was found that reducing the layer spacing and altering the functional groups can further optimize the spectral selectivity of Ti₃C₂T_x MXene. In addition to further improving the thermal insulation of the substrate (which could be achieved, for instance, through aerogel or fiber skeleton filling^6,37), these strategies could further enhance the multispectral-compatible camouflage performance. These speculations should be verified through further research.

Experimental methods

Fabrication of GS structured MAMMs

Details about the materials, experimental procedures, and characterization methods are provided in the ESI.† Additional characterizations of the SiOC used in this work, the orientation optimization of the GS architecture, and the properties of the GGS microwave blackbody architecture have been reported in references number 23, 15 and 7, respectively.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 52072306, U22A20129, and 52102122), the National Key R&D Program of China (2021YFB3701500), the National Defense Basic Scientific Research Program of China (JCKYS2019607001), and the Fundamental Research Funds for the Central Universities (3102019PJ008 and 3102018jcc002). The authors would like to express their gratitude to Northwestern Polytechnical University's Analytical & Testing Center for the SEM and TEM tests.

Notes and references

1. Y. Huang, Y. N. Zhu, B. Qin, Y. W. Zhou, R. Qin, P. Ghosh, M. Qiu and Q. Li, Nanophotonics, 2022, 11, 3613–3622.
2. H. Zhu, Q. Li, C. Tao, Y. Hong, Z. Xu, W. Shen, S. Kaur, P. Ghosh and M. Qiu, Nat. Commun., 2021, 12, 1805.
3. T. Kim, J. Y. Bae, N. Lee and H. H. Cho, Adv. Funct. Mater., 2019, 29, 1807319.
4. B. Liu, Z. Chen, Z. Li, J. Shi, D. Zhao, L. Liu and H. Wang, Opt. Eng., 2020, 59, 127107.
5. C. L. Ding, X. S. Deng, D. P. Zhao, Z. S. Chen, X. Y. Lv, Z. G. Li and J. M. Shi, Opt. Mater., 2022, 131, 112622.
6. H. Zhu, Q. Li, C. Zheng, Y. Hong, Z. Xu, H. Wang, W. Shen, S. Kaur, P. Ghosh and M. Qiu, Light Sci. Appl., 2020, 9, 60.
7. L. Yao, S. Zhou, L. Pan, H. Mei, Y. Li, K. G. Dassios, P. Colombo, L. Cheng and L. Zhang, Adv. Funct. Mater., 2022, 33.
8. M. Y. Pan, Y. Huang, Q. Li, H. Luo, H. Z. Zhu, S. Kaur and M. Qiu, Nano Energy, 2020, 69.
9. Z. X. Wang, Y. Z. Cheng, Y. Nie, X. Wang and R. Z. Gong, J. Appl. Phys., 2014, 116, 054905.
10. S. M. Zhong, W. Jiang, P. P. Xu, T. J. Liu, J. F. Huang and Y. G. Ma, Appl. Phys. Lett., 2017, 110.
11. M. L. Li, B. Muneer, Z. X. Yi and Q. Zhu, Adv. Opt. Mater., 2018, 6, 1701238.
12. M. S. Ergoktas, G. Bakan, E. Kovalska, L. W. Le Fevre, R. P. Fields, P. Steiner, X. Yu, O. Salihoglu, S. Balci, V. I. Fal'ko, K. Novoselov, R. A. W. Dryfe and C. Kocabas, Nat. Photonics, 2021, 15, 493–498.
13. R. Zhu, Z. Zhang, J. Wang, C. Xu, S. Sui, X. Wang, T. Liu, Y. Zhu, L. Zhang, J. Wang and S. Qu, Opt. Express, 2021, 29, 20150–20159.
14. C. Zhang, X. Wu, C. Huang, J. Peng, C. Ji, J. Yang, Y. Huang, Y. Guo and X. Luo, Adv. Mater. Technol., 2019, 4, 1900063.
15. L. Yao, W. Q. Yang, S. X. Zhou, H. Mei, Y. Li, K. G. Dassios, R. Riedel, C. D. Liu, L. F. Cheng and L. T. Zhang, Acta Mater., 2023, 249, 118803.
16. L. Li, M. K. Shi, X. Y. Liu, X. X. Jin, Y. X. Cao, Y. Y. Yang, W. J. Wang and J. F. Wang, Adv. Funct. Mater., 2021, 31, 2101381.
17. Y. Li, C. Xiong, H. Huang, X. Peng, D. Mei, M. Li, G. Liu, M. Wu, T. Zhao and B. Huang, Adv. Mater., 2021, 33, e2103054.
18. J. Lipton, J. A. Rohr, V. Dang, A. Goad, K. Maleski, F. Lavini, M. K. Han, E. H. R. Tsai, G. M. Weng, J. M. Kong, E. Riedo, Y. Gogotsi and A. D. Taylor, Matter, 2020, 3, 546–557.
19. M. Ghidiu, M. R. Lukatskaya, M. Q. Zhao, Y. Gogotsi and M. W. Barsoum, Nature, 2014, 516, 78–81.
20. A. Sarycheva, A. Polemi, Y. Liu, K. Dandekar, B. Anasori and Y. Gogotsi, Sci. Adv., 2018, 4, eaau0920.
21. Q. Jiang, N. Kurra, K. Maleski, Y. J. Lei, H. F. Liang, Y. Z. Zhang, Y. Gogotsi and H. N. Alshareef, Adv. Energy Mater., 2019, 9, 1901061.
22. Y. Luo, K. Kikuta, Z. L. Han, T. Takahashi, A. Hirose and H. Toshiyoshi, Electron. Lett., 2015, 51, 733–734.
23. L. Yao, W. Q. Yang, S. X. Zhou, H. Mei, L. F. Cheng and L. T. Zhang, Carbon, 2022, 196, 961–971.
24. F. Shahzad, M. Alhabeb, C. B. Hatter, B. Anasori, S. Man Hong, C. M. Koo and Y. Gogotsi, Science, 2016, 353, 1137–1140.
25. S. A. Shah, T. Habib, H. Gao, P. Gao, W. Sun, M. J. Green and M. Radovic, Chem. Commun., 2016, 53, 400–403.
26. L. Wang, P. Song, C. T. Lin, J. Kong and J. Gu, Research, 2020, 4093732.
27. J. Liu, T. Chen, P. Juan, W. Peng, Y. Li, F. Zhang and X. Fan, ChemSusChem, 2018, 11, 3758–3765.
28. X. Gao, B. Wang, K. Wang, S. Xu, S. Liu, X. Liu, Z. Jia and G. Wu, J. Colloid Interface Sci., 2021, 583, 510–521.
29. G. Chen, H. Wang, X. Wei, Y. Wu, W. Gu, L. Hu, D. Xu and C. Zhu, Sens. Actuators, B, 2020, 312, 127951.
30. Y. Fang, Z. Liu, J. Han, Z. Jin, Y. Han, F. Wang, Y. Niu, Y. Wu and Y. Xu, Adv. Energy Mater., 2019, 9, 1803406.
31. M. Naguib, V. N. Mochalin, M. W. Barsoum and Y. Gogotsi, Adv. Mater., 2014, 26, 992–1005.
32. H. J. Duan, H. X. Zhu, J. F. Gao, D. X. Yan, K. Dai, Y. Q. Yang, G. Z. Zhao, Y. Q. Liu and Z. M. Li, J. Mater. Chem. A, 2020, 8, 9146–9159.
33. F. Zhang, C. Li, Y. Fan, R. Yang, N. H. Shen, Q. Fu, W. Zhang, Q. Zhao, J. Zhou, T. Koschny and C. M. Soukoulis, Adv. Mater., 2019, 31, e1903206.
34. M. A. Seo, H. R. Park, S. M. Koo, D. J. Park, J. H. Kang, O. K. Suwal, S. S. Choi, P. C. M. Planken, G. S. Park, N. K. Park, Q. H. Park and D. S. Kim, Nat. Photonics, 2009, 3, 152–156.
35. A. Shahsafi, P. Roney, Y. Zhou, Z. Zhang, Y. Xiao, C. Wan, R. Wambold, J. Salman, Z. Yu, J. Li, J. T. Sadowski, R. Comin, S. Ramanathan and M. A. Kats, Proc. Natl. Acad. Sci. U. S. A., 2019, 116, 26402–26406.
36. D. G. Baranov, Y. Xiao, I. A. Nechepurenko, A. Krasnok, A. Alu and M. A. Kats, Nat. Mater., 2019, 18, 920–930.
37. Z. M. An, Y. P. Li, X. G. Luo, Y. X. Huang, R. B. Zhang and D. N. Fang, Matter, 2022, 5, 1937–1952.
38. X. Deng, Q. Nie, Y. Wu, H. Fang, P. Zhang and Y. Xie, ACS Appl. Mater. Interfaces, 2020, 12, 26200–26212.
39. J. H. Zhao, Y. Wang, G. F. Ding, Y. N. Sun and G. L. Wang, J. Micromech. Microeng., 2014, 24, 115013.
40. D. W. Abueidda, R. K. Abu Al-Rub, A. S. Dalaq, D. W. Lee, K. A. Khan and I. Jasiuk, Mech. Mater., 2016, 95, 102–115.
41. A. Mirabolghasemi, A. H. Akbarzadeh, D. Rodrigue and D. Therriault, Acta Mater., 2019, 174, 61–80.
42. Q. B. Wen, Z. J. Yu and R. Riedel, Prog. Mater. Sci., 2020, 109, 100623.
43. Z. Meng, C. H. Tian, C. L. Xu, J. F. Wang, S. N. Huang, X. H. Li, B. Y. Yang, Q. Fan and S. B. Qu, Infrared Phys. Technol., 2020, 110, 103469.
44. G. D. Sorarù, K. Girardini, M. Narisawa and M. Biesuz, J. Am. Ceram. Soc., 2021, 104, 3097–3104.
45. S. X. Zhou, L. Yao, T. Zhao, H. Mei, L. F. Cheng and L. T. Zhang, Carbon, 2022, 196, 253–263.
46. J. Li, X. T. Yuan, C. Lin, Y. Q. Yang, L. Xu, X. Du, J. L. Xie, J. H. Lin and J. L. Sun, Adv. Energy Mater., 2017, 7.
47. N. P. Avdelidis and A. Moropoulou, Energy Build., 2003, 35, 663–667.
48. L. Long, J. X. Xu, H. Luo, P. Xiao, W. Zhou and Y. Li, J. Am. Ceram. Soc., 2020, 103, 6869–6880.
49. W. Zhou, L. Long and Y. Li, J. Mater. Sci. Technol., 2019, 35, 2809–2813.
50. K. Wu, X. Zhao, T. G. Bifano, S. W. Anderson and X. Zhang, Adv. Mater., 2022, 34, e2109032.
51. Q. P. Liu, H. L. Xing, H. Wang and X. L. Ji, J. Mater. Sci.: Mater. Electron., 2019, 30, 19760–19769.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3mh00611e

---