Iron-catalyzed laser-induced graphene on cellulose paper for solar-driven interfacial evaporation

Abstract

Solar-driven interfacial evaporation has emerged as a promising strategy to produce freshwater via seawater desalination. While cellulose-based photothermal materials have garnered significant attention in solar steam generation, conventional surface modification techniques (e.g., coating and carbonization) often suffer from cumbersome preparation and limited design flexibility. This study pioneers a non-destructive laser-processing strategy to simultaneously induce spatially controlled graphene domains and iron oxide nanostructures on chemically modified cellulose matrices. Through synergistic integration of broadband light absorption (∼93.28% solar-weighted absorptance) with hierarchical water channels, the two-dimensional evaporator achieves exceptional evaporation performance (1.62 kg m⁻² h⁻¹, 99.0% efficiency) under 1 sun irradiation. Remarkably, the evaporation performance increases to 1.82 kg m⁻² h⁻¹ when configured into a three-dimensional architecture via simple single folding. Detailed characterizations reveal that laser-induced carbothermal reduction generates iron oxide–graphene composites as light absorbers, while preserving cellulose's inherent hydrophilicity for rapid capillary pumping. Notably, the engineered architecture demonstrates enhanced mechanical robustness (234% improvement in tensile strength) and programmable foldability, expanding applicability across diverse desalination scenarios. This laser-direct-writing paradigm establishes a sustainable pathway for developing next-generation cellulose-based solar evaporators.

---

New concepts

In this manuscript, the concept of laser-induced graphene has been proven to be directly applicable in the fabrication of cellulose paper-based solar interfacial evaporators. Although existing evaporators can achieve the secondary transfer of laser-induced graphene onto paper-based evaporators, the process flow is cumbersome, large-scale preparation is costly, long-term stability is weak, and the method lacks the convenience of direct laser engraving on paper substrates. This research emphasizes that cellulose paper with good moldability, after chemical treatment with a tannic acid–iron solution, can be directly laser-induced into graphene to obtain an excellent interfacial evaporation photothermal material while achieving a multi-fold increase in the mechanical strength. Simultaneously, the synergy between graphene and iron oxide in the composite layer significantly enhances the photothermal conversion efficiency, realizing ultra-efficient localized water evaporation in solar interfacial evaporation. This study covers the complete research trajectory from low-cost, readily available cellulose paper to the final product becoming a photothermal material for solar interfacial evaporation. Therefore, it proves the proof-of-concept of applying reasonable processing and design to improve the evaporation rate and photothermal efficiency. The convenience and importance of combining specific cellulose paper-based materials have been proven; this, combined with the latest advances in laser-induced graphene and 3D evaporator design, represents an interesting research area.

1. Introduction

The intensifying freshwater scarcity crisis, compounded by population growth, industrialization, and climate change, has emerged as one of the most pressing global challenges of the 21st century,^1,2 with over 2 billion people facing severe water stress despite 71% of Earth's surface being water-covered yet only <1% directly usable.^2–5 This urgency has propelled advancements in sustainable desalination technologies, where conventional reverse osmosis faces critical limitations including high energy consumption (>3 kW h m⁻³) and substantial carbon footprints.^6–10 In contrast, solar-driven interfacial evaporation (SDIE) has emerged as a transformative solution, achieving significant evaporation rates through localized heat generation at water–air interfaces.^11–16 Overall, the evaporation performance of SDIE^17,18 has been largely improved through rational design of photothermal materials that synergistically optimize solar absorption, thermal management, and water transport. While carbon-based materials (e.g., reduced graphene oxide and carbon nanotubes) initially dominated due to their broadband light absorption and chemical stability, practical deployment is hindered by scalability challenges (150–500 USD per m² production costs) and sustainability concerns.¹⁹ Recent breakthroughs identify cellulose paper as an excellent substrate for SDIE, leveraging its hierarchical porosity (10–100 μm pore channels) for capillary-driven water transport,²⁰ hydroxyl-rich surfaces for nanoparticle anchoring,^21–23 and ultralow thermal conductivity (0.04–0.06 W m⁻¹ K⁻¹) to suppress heat dissipation. Unlike natural wood requiring complex delignification, commercial cellulose paper enables facile structural engineering (cutting, folding, and laser processing) to tailor wettability and mechanical flexibility, positioning it as a scalable platform for next-generation SDIE systems.²⁴

Recent advancements in cellulose paper-based solar interfacial evaporators have demonstrated enhanced evaporation rates through chemical modification strategies. Surface engineering of cellulose papers via chemical treatment optimizes photothermal conversion efficiency while improving processability. Pioneering work by Liu et al.²⁵ utilized citric acid reduction to deposit gold nanoparticles on air-laid paper, achieving 87% solar absorption (400–800 nm) and validating cellulose composites as viable photothermal components. Subsequent developments include MXene/cellulose membranes fabricated through MXene suspension immersion and vacuum drying, exhibiting 94.1% broadband absorption (300–1500 nm) under wet conditions with a 1.44 kg m⁻² h⁻¹ evaporation rate under 1 sun irradiation.²⁶ Qin's²⁷ group engineered a bilayer evaporator featuring Fe₃O₄-embedded cellulose fibers as the photothermal layer, delivering a 1.22 kg m⁻² h⁻¹ evaporation rate. The intrinsic nanofibrillar architecture of mass-producible cellulose paper (CP) enables controlled growth of plasmonic nanoparticles, exemplified by Dong et al.'s Ag-CP composite achieving 93.7% solar absorption (250–2500 nm) and 85.2% thermal efficiency. Despite these advances, there is an urgent need to develop facile fabrication techniques for converting cellulose paper substrates into porous photothermal materials suitable for direct application in SDIE.

Inspired by laser engraving technology and the intrinsic hydrophilicity of cellulose paper, we engineered a high-performance cellulose–graphene composite by coating the substrate with a tannic acid–iron complex, followed by laser-induced carbonization. This non-destructive laser processing simultaneously generated a photothermally active carbonized layer (laser-induced graphene, LIG) decorated with iron oxide nanoparticles. The LIG architecture, characterized by extended π–π* conjugated systems, achieves broadband solar absorption (93.28% solar-weighted absorptance) with exceptional near-infrared response (>90% at 800–1200 nm). Concurrently, the laser treatment revealed the cellulose's inherent 3D interconnected microfluidic network (channel width: 10–200 μm), enabling dual functionality: (1) enhanced thermal localization at the water–air interface (axial temperature gradient: 28.4 °C mm⁻¹) through minimized conductive/convective/radiative losses, and (2) rapid capillary-driven water transport. The composite's hierarchical porosity and high specific surface area synergistically facilitated vapor diffusion, achieving an evaporation rate of 1.62 kg m⁻² h⁻¹ with 99.0% solar-to-vapor efficiency under 1 sun irradiation.²⁸ Compared with other laser-induced graphene methods, such as those using different substrates^29–31 and those employing flame retardant-treated paper,^32,33 the method proposed in this study exhibits the advantages of low cost, high production efficiency, excellent evaporation performance, and environmental friendliness. This laser-induced graphene strategy demonstrates exceptional scalability and salt-resistant operation, establishing a universal paradigm for sustainable desalination and wastewater remediation technologies.

2. Experimental section

2.1. Materials

Cellulose papers (thickness of 0.17 mm, density of 0.08 kg m⁻³, fiber-related pore size ranging from 10 to 12 μm, fiber width ≤ 50 μm, and porosity of 41.94%.) were purchased from Alibaba Company. Tannic acid (AR, C₇₆H₅₂O₄₆) was obtained from Shanghai Macklin Biochemical Technology Co., Ltd. Acacia (powder, pharmaceutical grade), ferric citrate pentahydrate (AR, FeC₆H₅O₇·5H₂O), and glycerol (AR, C₃H₈O₃) were obtained from Aladdin Chemical Co., Ltd. Deionized water was provided by Putian Yuxin Water Treatment Equipment Co., Ltd. All chemical reagents were directly utilized as received without additional purification.

2.2. Fabrication of cellulose paper-based evaporators

All cellulose paper (CP) samples were sectioned into 2 × 2 cm² sheets for evaporation testing. Laser carbonization was performed using a 10.6 μm continuous-wave CO₂ laser system (XB4060, Liaocheng First Laser Technology, 100 W) with optimized parameters: 10% laser power ratio, 0.4 mm² spot size, 100 mm s⁻¹ scanning speed, and 0.01 mm scanning interval. Cellulose paper/laser (CP/L) samples were fabricated via direct laser engraving of pristine CP. For the preparation of cellulose paper/Fe (CP/Fe), a tannic acid–iron coordination complex solution was prepared by dissolving 33 g of tannic acid in 72 g of deionized water at 60 °C under constant magnetic stirring (500 rpm) for 30 min. Subsequently, 5 g of gum arabic, 8 g of glycerol, and 7 g of ferric ammonium citrate pentahydrate were sequentially added to the solution while maintaining the temperature at 60 °C, with continuous stirring until complete dissolution was achieved. The resulting solution was uniformly applied using a brush onto CP substrates and air-dried under ambient conditions. Cellulose paper/Fe-laser (CP/Fe-L) composites were prepared by coating CP with the tannic acid–iron solution and curing at 20 °C/60% relative humidity for 12 h, followed by laser engraving under identical parameters.

2.3. Solar steam generation experiment

During laboratory-scale evaporation testing, paper-based evaporators were assembled with expanded polyethylene foam insulation. Solar steam generation experiments were conducted using a xenon lamp (PLS-SXE300, Beijing PerfectLight Co., Ltd) calibrated to 1 sun intensity (1 kW m⁻²) using a spectroradiometer (SM206 plus, Shanghai Xinbao Technology Co., Ltd). The experiments were typically conducted at 25 °C and 45 %RH. Mass change monitoring was conducted using a high-precision analytical balance (PR224ZH/E, Ohaus Instruments (Shanghai) Co., Ltd) with 0.1 mg resolution, sampling at 10-s intervals. The surface temperatures of the developed solar evaporators were quantitatively measured using an infrared thermal camera (HIKMICRO H16). T-type thermocouples (Omega Engineering, ±0.1 °C accuracy) were used for air/water temperature measurements. The solar-to-steam conversion efficiency (η) was calculated as:³⁴

Q = c(T₂ − T₁) (3)

where ṁ is the water evaporation rate per unit area with the dark evaporation rate subtracted. h_LV is the total enthalpy of the liquid–vapor phase change including sensible heat and liquid-to-vapor latent heat of water, T is the temperature of vaporization, C_opt is the optical concentration of solar energy, q_sol is the nominal direct solar irradiation under 1 sun irradiation, Q is the sensible heat of water of unit mass, c is the specific heat of water, which can be assumed as a constant (4.2 kJ kg⁻¹ K⁻¹), T₂ is the temperature of vaporization and T₁ is the initial temperature of water.

2.4. Characterizations

The surface morphology of four specimen types was characterized via scanning electron microscopy (SEM, TESCAN MIRA LMS) under 5 kV acceleration voltage with 5 nm Au/Pd sputter coating. X-ray diffraction (XRD) analysis was conducted using a Rigaku Ultima IV diffractometer with divergent-beam optics: Cu Kα radiation (λ = 1.5406 Å, 40 kV/40 mA) was employed for CP and CP/L samples across 2θ = 5–90° (step size: 0.02°, scan rate: 2° s⁻¹), while Co Kα (λ = 1.7902 Å) was utilized for CP/Fe and CP/Fe-L to mitigate fluorescence artifacts. XPS measurements were conducted via a scanning XPS microprobe (Nexsa, Thermo Fisher Scientific) equipped with Al-Kα radiation. Fourier-transform infrared spectroscopy (FTIR, Thermo Fisher Scientific Nicolet iS5) was conducted in attenuated total reflectance mode (Ge crystal, 45° incidence) with 4 cm⁻¹ resolution across 400–4000 cm⁻¹. Raman spectra were acquired using a Horiba LabRAM HR Evolution system equipped with a 532 nm diode laser (50 mW, 100× objective, 1800 g mm⁻¹ grating), integrating 10 accumulations at 10 s exposure. Porosity measurements were performed using a high-performance automated mercury intrusion porosimeter (Mercury intrusion porosimetry, Micromeritics AutoPore IV 9520). Optical reflectance was quantified via UV-vis-NIR spectroscopy (PerkinElmer Lambda950) with a 150 mm integrating sphere, referencing BaSO₄ baseline from 250 to 2500 nm. Total solar absorptance (α)³⁵ was calculated as:where I(λ) and R(λ) are the sunlight intensity and light reflectance at wavelength λ. The CP/Fe-L composite exhibited zero optical transmittance across 250–2500 nm (ASTM E903-22), confirming complete photon capture within the laser-engineered graphene matrix. Pre- and post-desalination seawater cation concentrations were quantified via inductively coupled plasma optical emission spectroscopy (ICP-OES, Agilent 720ES) following ASTM D1976-20 protocols. Samples were acidified to pH < 2 with ultrapure HNO₃ (TraceSELECT®) and filtered through 0.22 μm nylon membranes prior to analysis, achieving method detection limits of 0.01 mg L⁻¹ for Na⁺, K⁺, Ca²⁺, and Mg²⁺. Dye degradation efficiency was evaluated using UV-vis spectroscopy (Shimadzu UV-3600i Plus) with 10 mm quartz cuvettes, measuring methyl orange (λ_max = 464 nm) and methylene blue (λ_max = 664 nm) absorbance before/after treatment. Spectral deconvolution employed the Kubelka–Munk function for diffuse reflectance correction (ISO 24560-1:2020), revealing >99.8% dye removal efficiency through combined photothermal degradation and adsorption mechanisms.

3. Results and discussion

As shown in Fig. 1(b), pristine cellulose paper (CP) exhibits a white appearance and super-hydrophilic properties (water contact angle ≈ 0°, 2.35 s fully immersed). SEM characterization reveals the microstructural features of the cellulose paper, where cellulose fibers are randomly interwoven through hydrogen bonding and van der Waals forces³⁶ to form a three-dimensional porous network structure with a broad pore size distribution ranging from approximately 10 to 50 μm. Notably, longitudinal grooves are observed on the fiber surfaces, which effectively enhance the specific surface area of the material (Fig. 1(e)). The inter-fiber junctions form micron-sized texture structures that facilitate light scattering and capillary-driven liquid transport. Direct CO₂ laser engraving applied to the pristine cellulose paper without pretreatment yields a material (denoted as CP/L, Fig. 1(c)) exhibiting yellowish-brown surface coloration. The laser-treated cellulose paper displays distinct ablation features, with the engraved regions demonstrating significant optical translucency.

Fig. 1 Fabrication process and SEM images of CP, CP/L, and CP/Fe-L. (a) Schematic diagram of the fabrication process of CP/Fe-L. CP/Fe-L was produced by applying laser treatment to cellulose paper after coating with tannic–iron solution. (b)–(d) Digital pictures of different samples: (b) CP, (c) CP/L, and (d) CP/Fe-L. (e)–(g) SEM images at different magnifications: (e) CP, (f) CP/L, and (g) CP/Fe-L.

The modified cellulose paper treated with a composite solution containing tannic acid, iron(III) citrate, glycerol, and gum arabic³⁷ (denoted as CP/Fe, Fig. S1c) exhibits a black surface coloration with near-zero optical transmittance. Microstructural analysis reveals that the deposited coating obscures the original fibrous network architecture, presenting a relatively smooth surface morphology (Fig. S1g) that compromises light-trapping capability through reduced scattering interfaces. Laser engraving of the CP/Fe composite (designated as CP/Fe-L, Fig. 1(a)) induces a sequential transformation process: the CO₂ laser irradiation first triggers thermal decomposition of the surface-bound iron–tannate complex into amorphous carbon matrices containing iron oxide nanoparticles, followed by their carbothermic reduction to iron carbide (Fe₃C). The resultant Fe₃C domains subsequently catalyze laser-induced graphene formation through localized reconstruction of sp²-hybridized carbon networks. The laser-processed cellulose paper develops multi-scale hierarchical textures comprising crisscrossed grooves and porous architectures on its surface. The observed groove patterns originate from the inherent fibrous network architecture of the original cellulose substrate. Microscopic characterization (Fig. S2) reveals co-existing micron- and nanoscale cavities within CP/Fe-L. The micrometer-sized pores form a hierarchical scattering matrix that facilitates multiple internal reflection via Mie scattering,^38,39 thereby extending the optical pathlength to several times the material's geometric thickness. Concurrently, nanoscale cavities induce Rayleigh scattering^40,41 through propagation direction randomization, particularly enhancing light-trapping efficiency in the ultraviolet-visible spectrum (300–800 nm), as evidenced by subsequent optical characterization. High-magnification SEM imaging further reveals that the micron-scale cavities are decorated with nanoparticulate metal oxide deposits. Subsequent spectroscopic analysis identified the predominant oxide phase as nonstoichiometric iron oxide (Fe_0.94O), with its crystallographic configuration being unambiguously confirmed through combined XRD and XPS characterization (Fig. 2(b) and (d)). Cross-sectional SEM imaging of CP/Fe-L (Fig. S3) demonstrates structural heterogeneity along the vertical axis: the basal region maintains the original cellulose fiber network, while the upper layer develops a microporous architecture with engineered structural hydrophilicity and thermal localization. This hierarchical porous architecture synergistically enhances hydraulic connectivity through capillary-driven water permeation, enabling efficient fluid transport from bulk water reservoirs to the photothermal interface. Moreover, the laser-induced microfluidic channels create two-dimensional water pathways that facilitate rapid salt ion diffusion, effectively suppressing salt accumulation through backflow and diffusion ion redistribution mechanisms.

Fig. 2 Characterization results of CP, CP/L, CP/Fe, and CP/Fe-L. (a) SEM elemental mapping of Fe, C, and O on CP/Fe-L, showing uniform distribution of Fe elemental on the surface. (b) XRD patterns and (c) Raman spectra of CP, CP/L, CP/Fe, and CP/Fe-L. (d) Fe 2p XPS spectra of CP/Fe-L. (e) XPS survey spectra of CP, CP/L, CP/Fe, and CP/Fe-L. (f) FTIR spectra of CP, CP/L, CP/Fe, and CP/Fe-L.

EDS mapping analysis confirms the homogeneous distribution of Fe, C, and O elements on the CP/Fe-L surface (Fig. 2(a)), indicating uniform coating of the tannic acid–iron (TA–Fe) solution on cellulose paper and subsequent formation of evenly dispersed iron oxides and graphitic phases during laser treatment. The XRD patterns of CP/Fe-L (Fig. 2(b)) exhibit characteristic peaks at 2θ = 26.5° (attributed to graphitic carbon^42,43) and additional peaks at 2θ = 36.07° (111), 41.90° (200), 60.74° (220), and 76.52° (222), corresponding to Fe_0.94O nanoparticles (JCPDS #077-2355).⁴⁴ These results confirm the oxidative conversion of the TA–Fe precursor into porous Fe_0.94O composites under laser-induced high-temperature conditions. Raman spectroscopy further verified the presence of graphitic layers (with characteristic D and G bands^45–47) formed through iron-catalyzed laser-induced graphene.

The X-ray photoelectron spectroscopy (XPS) survey spectra acquired for the CP, CP/L, CP/Fe, and CP/Fe-L samples (Fig. 2(d) and (e)) predominantly exhibited carbon and oxygen as the major elements, with a notably weak but distinct Fe signal detected in the CP/Fe-L spectrum, thereby confirming the successful incorporation of iron species into the composite material.⁴³ A detailed high-resolution analysis of the Fe 2p region (Fig. 2(d)) revealed two sets of characteristic peaks corresponding to different oxidation states: the peaks located at 710.19 eV and 722.99 eV were assigned to Fe²⁺ in the Fe 2p_3/2 and Fe 2p_1/2 orbitals, respectively, while those at 712.19 eV and 725.14 eV were attributed to Fe³⁺ species, accompanied by satellite features distributed around both the Fe 2p_3/2 and Fe 2p_1/2 regions that are typical of iron oxide systems. Of particular interest was the observation of a low-intensity “pre-peak” near 706 eV,⁴⁸ which has been interpreted as arising from Fe ions exhibiting oxidation states lower than their nominal values, a phenomenon potentially linked to the generation of structural defects—such as oxygen vacancies or lattice distortions—during sample preparation processes conducted under vacuum conditions. These XPS findings were further corroborated by XRD analysis, which identified the cubic NaCl-type crystal structure characteristics of wüstite (Fe_0.94O), where the oxygen sublattice remains fully occupied, while significant vacancies exist at iron sites, thereby indicating an oxygen-excess composition that formally corresponds to iron deficiency within the crystalline framework.

FTIR spectroscopy (Fig. 2(f)) revealed a reduction in the characteristic peaks of cellulose paper (CP) after chelation treatment. Absorption bands between 1800 and 400 cm⁻¹, attributed to cellulose in pristine CP, included distinct peaks at 3329 cm⁻¹ (–OH stretching vibration), 2908 cm⁻¹ (C–H stretching), 1641 cm⁻¹ (H–O–H bending), and 1026 cm⁻¹ (C–O–C skeletal vibration).^49–51 The broad –OH stretching peak at 3329 cm⁻¹ shifted to 3216 cm⁻¹ in the composite film, indicating interactions between iron oxides and cellulose hydroxyl groups. Laser engraving preserved the characteristic cellulose peaks (e.g., 1026 cm⁻¹), confirming retained hydrophilicity critical for efficient water transport during evaporation.

The light absorption characteristics of the CP/Fe-L evaporator across the broad spectral range of 280–2500 nm were further investigated. As shown in Fig. 3(a), (b) and Fig. S2, the CP/Fe-L evaporator exhibited a significantly enhanced total light absorption rate (93.28%) compared to CP (29.67%), CP/L (53.44%) and CP/Fe (86.15%), attributed to its superior light-trapping architecture, while demonstrating near-zero optical transmittance (Fig. S4). This high solar absorption efficiency ensures effective utilization of incident sunlight during vapor generation. The photothermal conversion capability of CP, CP/L, CP/Fe, and CP/Fe-L evaporators was evaluated by monitoring their surface temperature evolution under 1 sun irradiation (Fig. 3(d)). Infrared thermal imaging before irradiation and at different exposure intervals (Fig. 3(e) and Fig. S6a) corroborated these findings. The CP/Fe-L surface temperature rapidly increased from 24.2 °C to 37.1 °C within the initial 30 s, ultimately reaching a steady-state temperature of 42.3 °C after 1 h of continuous irradiation. This final temperature substantially exceeded those of CP (33.7 °C) and CP/L (37.6 °C) surfaces and slightly surpassed that of CP/Fe (41.9 °C). CP/Fe-L demonstrates superior water absorption (Fig. S5), a higher evaporation rate, and efficient heat dissipation, ultimately resulting in minimal temperature difference compared to CP/Fe. These results demonstrate the CP/Fe-L evaporator's significant photothermal conversion capability, attributed to its excellent light absorption performance.

Fig. 3 Photothermal conversion and water wettability of CP, CP/L, CP/Fe, and CP/Fe-L. (a) Reflectance and (b) absorptance spectra of CP, CP/L, CP/Fe, and CP/Fe-L, along with the irradiance density distribution of the normalized tilted sunlight spectrum at air mass 1.5G (AM 1.5G). (c) Water droplets on CP and CP/Fe-L, along with their corresponding water contact angles. (d) Surface temperature of CP, CP/L, CP/Fe, and CP/Fe-L during the evaporation process under 1 sun irradiation. (e) Infrared image of CP/Fe-L in 1 hour under 1 sun irradiation (at 0 s, 5 s, 10 s, 20 s, 30 s, and 3600 s, respectively).

Further measurements of contact angles for different samples (Fig. 3(c) and Fig. S5) revealed distinct wetting behaviors. The CP, CP/L, and CP/Fe-L achieved complete water wetting within 2.35 s, 1.83 s, and 0.51 s, respectively, while the CP/Fe evaporator exhibited a contact angle of 73°, indicating reduced hydrophilicity compared to CP/Fe-L. To systematically evaluate the influence of iron content, samples with varying ferric citrate pentahydrate (FCP) mass fractions were analyzed: CP/Fe-L-0.6 (2.5% FCP), CP/Fe-L-1.0 (4.1%), CP/Fe-L-1.4 (5.6%), CP/Fe-L-1.8 (7.1%, denoted as CP/Fe-L), and CP/Fe-L-2.2 (8.5%). With the exception of CP/Fe-L-2.2 (wetting time: 0.99 s), samples CP/Fe-L-0.5, CP/Fe-L-1.0, and CP/Fe-L-1.4 displayed contact angles of 44°, 37°, and 31° at 10 s, respectively. CP/Fe-L-1.8 (CP/Fe-L) achieves complete wetting within 0.51 s, demonstrating the fastest infiltration rate and confirming optimal hydrophilicity, attributed to its balanced iron-loading and surface energy modulation. To further elucidate capillary dynamics,^52,53 infrared imaging was used to capture evaporation-driven water uptake (Fig. S7a) and vertical capillary rise (Fig. S7b) behaviors. The CP/Fe-L evaporator demonstrated rapid capillary-driven water transport, achieving a 3 cm vertical rise within 88 s and full infiltration of a 2 × 2 cm² area in 144 s. Calculated water transport rates (0.25 mm s⁻¹, Fig. S7c) validate its superior hydraulic conductivity, ensuring sustained water supply during vapor generation.

It is hypothesized that carbonization increases the substrate porosity to facilitate rapid wetting. Mercury intrusion porosimetry measurement showed that the porosity of CP is 41.94%, while that of CP/Fe-L is 44.47%, indicating a slight increase. Capillary pumping tests in Fig. S8 demonstrate that due to cellulose's inherently higher water affinity,⁵⁴ evidenced by its 2.23 times greater absorption capacity compared to carbon materials in Figs. S7d and S10, CP exhibits superior capillary pumping capability over CP/Fe-L at extended timescales (≥5 min), ultimately achieving greater water rise heights. Additionally, both Fig. 3(c) and Fig. S8 reveal CP/Fe-L's accelerated wetting behavior where the contact angle reaches zero within 0.51 s. The initial rapid uptake stems from CP/Fe-L's enhanced initial wetting (Fig. 3(c)) and increased porosity, enabling superior short-distance capillary velocity. However, this advantage proves unsustainable as Fig. S8 shows the capillary rise rate (derived from the slope) progressively falls below that of CP after 10 s, causing CP to surpass CP/Fe-L's water height by approximately 5 min. Permeability observations from capillary rise (Fig. S7) and liquid transport experiments (Fig. S11) confirm complete wetting across the 20 mm × 20 mm CP/Fe-L evaporator within 144 s. Furthermore, the asymmetrical substrate structure enables combined horizontal and vertical liquid transport toward the graphene/iron oxide nanocomposite layer, benefiting thin-film evaporation as supported by evaporation rate comparisons in Fig. 4(a). Laser treatment increases substrate porosity from 41.94% to 44.47%, confirming its role in promoting rapid wetting. This hierarchical pore structure, combining macrochannels for rapid bulk flow and nanochannels for surface wetting, minimizes vaporization enthalpy penalties while maintaining thermal localization efficiency.

Fig. 4 Indoor evaporation performance of CP, CP/L, CP/Fe, and CP/Fe-L. (a) Mass change curves of deionized water for CP, CP/L, CP/Fe, and CP/Fe-L under 1 sun irradiation. (b) Mass change curves of deionized water for CP, CP/L, CP/Fe, and CP/Fe-L in dark condition. (c) Evaporation efficiency of CP, CP/L, CP/Fe, and CP/Fe-L. (d) Evaporation rate and surface temperature of the CP/Fe-L evaporator under 1 sun irradiation. (e) Mass change curves of CP/Fe-L under 0.5 sun, 1 sun, and 2 sun irradiation, and (f) in seawater with different salt concentrations under 1 sun irradiation (including pure water, 2 wt%, 3.5 wt%, 7 wt%, and 10 wt%).

The solar-driven steam generation performance of the CP/Fe-L evaporator was evaluated by monitoring mass loss of water under 1 sun irradiation. Evaporation rate experiments under 1 sun irradiation (Fig. S9) demonstrated a non-linear relationship between iron concentration and evaporation performance. While increasing FCP mass fraction enhanced evaporation rates initially, a maximum value of 1.62 kg m⁻² h⁻¹ was observed at 7.1% FCP (CP/Fe-L-1.8), beyond which performance plateaued due to pore blockage and reduced capillary transport efficiency. This trend aligns with hydrophilicity data, confirming that excessive iron loading compromises structural integrity despite improved light absorption. Based on this synergistic evaluation, CP/Fe-L-1.8 was selected as the optimal solar interfacial evaporator. Evaporation rates were calculated from the slope of mass-versus-time curves (Fig. 4(a)). As shown in Fig. 4(d), the CP/Fe-L evaporator exhibited rapid evaporation rate escalation within the first 10 min, followed by stabilization under prolonged irradiation, a trend consistent with its surface temperature dynamics. Comparative analysis revealed that CP/Fe-L achieved a significantly higher mass loss rate than CP, CP/L, and CP/Fe under identical conditions (Fig. 4(a)). In contrast, dark-state evaporation experiments (Fig. 4(b)) demonstrated minimal performance differences among all samples. Specifically, the evaporation rates of the CP, CP/L, CP/Fe, and CP/Fe-L evaporators within one hour were measured as 0.49, 0.92, 1.29, and 1.62 kg m⁻² h⁻¹ under 1 sun irradiation, respectively (Fig. S13). Using eqn (1) to subtract dark-state evaporation contributions, the net solar-to-vapor efficiencies within one hour were calculated as 22.7% (CP), 51.3% (CP/L), 75.7% (CP/Fe), and 99.0% (CP/Fe-L). The high efficiency of CP/Fe-L demonstrates its exceptional interfacial evaporation activity, resulting from the synergistic enhancement of broadband light absorption (>93% in 280–2500 nm) and photothermal conversion capability. To optimize substrate selection, three cellulose paper grades (high-speed, medium-speed, and low-speed) were comparatively assessed for hydraulic properties (Fig. S10). High-speed cellulose paper demonstrated superior water absorption kinetics (capillary rise rate: 0.33 mm s⁻¹) and total absorption capacity (3.2 mL cm⁻²), justifying its selection as the evaporator substrate. The outstanding evaporation performance of CP/Fe-L thus derives from both engineered photothermal layers and the substrate's intrinsic water transport superiority, enabling sustained evaporation without salt accumulation or hydraulic limitations.

However, natural sunlight intensity under outdoor conditions is inherently variable. In this context, the influence of light intensity on the evaporation rate and efficiency of CP/Fe-L was systematically investigated (Fig. 4(e)). The surface temperature of CP/Fe-L increased significantly with irradiation intensity escalation from 0.5 sun to 2 sun, as evidenced by stabilized surface temperatures after 1 h irradiation increasing from 35.1 °C (0.5 sun) to 42.3 °C (1 sun) and 54.0 °C (2 sun) (Fig. S12). Correspondingly, the evaporation rate increased from 0.72 kg m⁻² h⁻¹ (0.5 sun) to 1.62 kg m⁻² h⁻¹ (1 sun) and 2.26 kg m⁻² h⁻¹ (2 sun). Notably, the evaporation rate exhibited a non-proportional enhancement with light intensity escalation, attributable to distinct rate-limiting factors under varying irradiance levels. Under low-intensity conditions (0.5 sun), the photon flux density primarily governed evaporation kinetics, whereas under high-intensity irradiation (2 sun),⁵⁵ amplified temperature gradients accelerated thermal dissipation to the ambient environment, thereby reducing the utilization efficiency of absorbed radiant energy. This mechanistic interpretation was corroborated by efficiency variations, peaking at 99.0% under 1 sun irradiation but declining to 78.0% (0.5 sun) and 70.3% (2 sun).

Comparative evaluation of CP/Fe-L's solar desalination performance across saline solutions with varying salinity levels (Fig. 4(f)) revealed a 31% reduction in the evaporation rate (1.12 kg m⁻² h⁻¹) at simulated seawater concentration (3.5 wt%) compared to optimal freshwater conditions. Despite this decline, CP/Fe-L maintained superior performance over CP (0.49 kg m⁻² h⁻¹) and CP/L (0.92 kg m⁻² h⁻¹) in freshwater, demonstrating robust desalination capability. The diminished mass loss rate in seawater relative to deionized water originated from reduced vapor pressure in saline solutions, which increases the energy barrier for water molecule evaporation.⁵⁶ Building on this, we conducted a 10-hour evaporation cycle under simulated seawater conditions in a laboratory setting (Fig. S19). Significant salt accumulation became evident on both sides of the evaporator after 4 hours. Over the ten hours, the rate of decline in the evaporation rate gradually stabilized, resulting in a final retention rate of 56%. This demonstrates its potential for sustained performance during long-term operation. Furthermore, the substrate's low cost and environmental friendliness make it suitable for scalable production and application in water evaporation.

The mechanical robustness of paper-based evaporators is critical for practical applications, as they must meet diverse operational demands. First, resistance to wave-induced impacts requires high mechanical strength and fatigue resistance to withstand dynamic stresses and stabilize the evaporation interface. Second, assembly processes demand materials with tailored flexibility and machinability to prevent microcracks or leakage caused by alignment errors, ensuring structural integrity. Third, frequent salt-flushing necessitates substrates with combined corrosion resistance and surface wear resistance to mitigate porosity alterations from salt-crystallization abrasion, thereby maintaining long-term evaporation efficiency. Fourth, foldable applications require stable microstructures after repeated bending to avoid fiber fracture or hydrophilic coating delamination, preserving evaporation rates over multiple cycles. The synergistic fulfillment of these properties ensures both mechanical durability and high evaporation efficiency in dynamic environments, enabling sustained desalination or water purification through stable hydraulic transport and photothermal conversion.

The mechanical performance of the cellulose paper substrate was enhanced through chemical treatment and laser engraving. The thickness variation after ferric tannate coating was negligible (Fig. S14c), allowing direct comparison of tensile fracture forces between CP and CP/Fe-L. As shown in Fig. S15, the tensile fracture force of CP/Fe-L (96.8 N) nearly doubled that of CP (50.8 N), with a corresponding tensile fracture stress of 58.08 MPa. After 3 h of continuous evaporation, CP/Fe-L retained 89% of its initial tensile strength (85.8 N vs. 96.8 N), demonstrating sufficient robustness for recyclability and long-term durability. The stress–strain curves of various samples in Fig. 5(a) reveal that compared to the untreated substrate (CP), laser engraving alone (CP/L) failed to improve tensile resistance and instead rendered the material more brittle. Mere coating with tannic acid–iron solution only provided a marginal enhancement (106%) in tensile fracture stress. Subsequent laser engraving of samples without (tannic-L) and with (CP/Fe-L) iron(III) citrate pentahydrate showed that CP/Fe-L achieved a 200% increase in tensile fracture stress relative to CP, while that of tannic-L reached 139%. This confirms that incorporating iron oxide nanoparticles significantly boosts mechanical performance compared to LIG alone. Furthermore, we quantified mechanical properties by calculating Young's modulus for both untreated (CP) and treated (CP/Fe-L) substrates before and after 3-hour evaporation. As presented in Table S1, CP's Young's modulus decreased from 330 MPa to 220 MPa post-evaporation, while CP/Fe-L maintained 830 MPa after evaporation from an initial 1050 MPa – representing a threefold stiffness improvement. Leveraging the inherent foldability of cellulose paper, macroscopic structural engineering was employed to enhance light trapping. As illustrated in Fig. 5(b) and Fig. S16, a simple V-shaped folding configuration increased the evaporation rate of CP/Fe-L to 1.82 kg m⁻² h⁻¹, representing a 12.35% enhancement over the flat design. This confirms the dual-scale (microstructural and macro-architectural) light confinement capabilities,⁵⁷ and converting a two-dimensional evaporator into a three-dimensional one through a single folding operation significantly increases the evaporation surface area while maintaining the same projected area as the two-dimensional evaporator. Meanwhile, it enables incident sunlight to undergo multiple reflections in the microgrooves, enhancing light absorption efficiency.⁵⁸ In addition, its structure can achieve continuous water transport from the bottom to the top through capillary action and transpiration effects, ensuring the water supply required for evaporation. Therefore, it exhibits more excellent evaporation performance, suggesting potential for advanced 3D evaporator designs. Moreover, we tested two additional 3D configurations (Fig. S17). Fold 1 exhibited a lower evaporation rate than the flat CP structure. This is attributed to restricted capillary-driven water transport, where gravity impedes upward flow, causing water supply lag and localized dry spot formation in the vaulted region, thereby reducing evaporation. In contrast, the concave structures at the base of both fold 2 and fold 3 remained wet after two hours. Combined with their effective light-trapping capability, these configurations enable superior light absorption and sustained water supply, resulting in enhanced evaporation performance. Long-term stability tests with deionized water (Fig. 5(d), (e) and Fig. S18) revealed superior performance retention for CP/Fe-L (88.6% after 540 min) compared to CP (85.4%), as evidenced by consistent mass loss curves and cyclic evaporation rates. This stability arises from the substrate's chemically reinforced matrix and defect-resistant photothermal layer, which minimize performance degradation under operational stresses.

Fig. 5 Mechanical properties, folding effect, evaporation stability and outdoor evaporation experiments of CP/Fe-L. (a) Tensile stress–strain curves of the five evaporator substrates. (b) Mass change curves of CP/Fe-L before and after folding under 1 sun irradiation. (c) Salt cation concentrations (Na⁺, Mg²⁺, K⁺, and Ca²⁺) of seawater before and after desalination using CP/Fe-L. Ten consecutive evaporation cycles with deionized water over a 10-hour period of CP/Fe-L: (d) mass change curves and (e) evaporation rate histograms. (f) Absorption spectra of methyl blue (+NaCl) and methyl orange (+NaCl) in the simulated dye wastewater and its condensed water. (g) Outdoor evaporation test performance in 3.5 wt% simulated salt water (location: 113°57′ E, 22°47′ N; date: May 2, 2025).

The concentrations of major cations (Na⁺, Mg²⁺, K⁺, and Ca²⁺) in seawater before and after desalination were analyzed via ICP-OES. As shown in Fig. 5(c), the reduction rates of cation concentrations post-treatment exceeded 96.1% compared to pre-distillation levels, with the distilled water fully complying with drinking water quality standards stipulated by the World Health Organization (WHO) and Environmental Protection Agency (EPA). To evaluate contaminant removal efficacy, methylene blue (MB, 20 mg L⁻¹) and methyl orange (MO, 20 mg L⁻¹)² were introduced as model dye pollutants in four simulated wastewater systems. The absorbance spectra of the treated distillate (Fig. 5(f), dashed lines) showed complete elimination of characteristic MB (590 nm) and MO (465 nm) peaks. Notably, in saline-containing control groups (MO + NaCl and MB + NaCl), the CP/Fe-L evaporator maintained robust dye removal performance, as evidenced by the absence of residual color in the collected distillates (Fig. S21), confirming its capability for simultaneous desalination and pollutant degradation.

For field validation, a 3.5 wt% saline solution was subjected to outdoor evaporation testing (location: 113°57′ E, 22°47′ N; date: May 2, 2025) under natural daylight irradiation (Fig. 5(g) and Fig. S22). Mass collection data recorded at 30-min intervals over 8 h (9:00–17:00) demonstrated stable performance, with a total evaporation output of 4.84 kg m⁻² despite variable cloud cover. This result underscores the CP/Fe-L evaporator's operational reliability under field conditions, achieving consistent freshwater production through adaptive light absorption and thermal management. In a humid environment, cellulose undergoes oxidation reactions, producing hydroperoxides as well as small-molecule organic acids such as formic acid and acetic acid. Transition metal ions can catalyze the decomposition of hydroperoxides through the Fenton reaction, generating reactive species like hydroxyl radicals, which further accelerate oxidative degradation.⁵⁹ However, cellulose paper can still maintain a relatively long lifespan under wet conditions.⁶⁰ Additionally, cellulose paper is low in cost and convenient to replace, so it can be replaced in a timely manner after the performance of the evaporator deteriorates, ensuring the continuity and high efficiency of seawater desalination.

4. Conclusion

In summary, we fabricated an efficient graphene/iron oxide composite solar evaporator using laser-induced graphene processing on cellulose paper, which features hierarchical micro–nano-porosity with uniformly anchored iron oxide nanoparticles and microparticles, achieving over 93.28% solar absorptivity from 280 to 2500 nm and intrinsic hydrophilicity. Under one-sun irradiation, the CP/Fe-L evaporator attains a 1.62 kg m⁻² h⁻¹ evaporation rate with 99.0% solar-to-vapor efficiency in its two-dimensional configuration, which increases to 1.82 kg m⁻² h⁻¹ when reconfigured into a three-dimensional architecture through single folding. The processed cellulose-based substrate exhibits enhanced mechanical strength with tripled stiffness, while enabling programmable foldability for designing customized 3D evaporators suitable for multisource water harvesting scenarios. This laser-engraving approach combines cost-effectiveness with scalable production, while the self-floating CP/Fe-L design demonstrates practical potential for large-scale seawater desalination across diverse environments due to its superior evaporation performance, environmental compatibility, mechanical strength, and structural reconfigurability.

Author contributions

Jing Zhang: manuscript reviewing and editing, original draft composition, visual design, result verification, research oversight, computational analysis, methodology development, formal assessment, data management, study conception. Yujun Wei: methodology implementation, experimental investigation. Penghui Chen: statistical analysis, data organization. Zhijian Huang: data collection and curation, conceptual design. Qixu Yu: project supervision, algorithm development. Jing Tan: software programming, methodology application. Hongzhen Zeng: visualization, validation. Shudong Yu: manuscript revision and editing, initial draft preparation, graphical visualization, data validation, resource provision, funding acquisition.

Conflicts of interest

The authors declare no conflicts of interest.

Data availability

The data supporting this article have been included as part of the SI. Additional figures supporting the main text, including SEM images, CP/Fe-L edge transmittance, contact angles of other samples, capillary rise experiments with substrates, and stress–strain tensile tests (mechanical properties), etc. See DOI: https://doi.org/10.1039/d5mh01166c

Acknowledgements

The authors would like to acknowledge the financial support received from National Key Research and Development Program of China (2024YFB3213600), Guangdong Basic and Applied Basic Research Foundation (2024A1515010854 and 2023A1515111194); National Natural Science Foundation of China (52405329 and 52475376); Shenzhen Science and Technology Program (szbo202317); One-Hundred Talents Program of Sun Yat-sen University.

References

1. T. Arunkumar, Y. Suh, H. W. Lim, S. Christopher and S. J. Lee, Desalination, 2024, 572, 117102.
2. J. Li, L. Mu, Q. Liu, Y. Zhang, R. Zhang, X. Zhu, C.-L. Sun, J. He and M. Qu, Nano Energy, 2024, 132, 110394.
3. G. Jing, X. Meng, W. Sun, P. B. Kowalczuk and Z. Gao, Environ. Sci. Water. Res., 2023, 9, 1290–1304.
4. M. Sheng, Y. Yang, X. Bin, S. Zhao, C. Pan, F. Nawaz and W. Que, Nano Energy, 2021, 89, 106468.
5. W. Li, Y. Wei, H. Zeng, Z. Huang, J. Wu and S. Yu, Energy Convers. Manage., 2024, 315, 118777.
6. S. J. Milani and G. Nabi Bidhendi, Water Sci. Eng., 2024, 17, 283–291.
7. Q. Zhang, Y. Zhang, Y. Shen, Q. Ye, Q. Cai and X. Wu, J. Environ. Chem. Eng., 2021, 9, 105787.
8. Y.-Q. Luo, F. Song, J.-M. Wu, F. Wang, X.-L. Wang and Y.-Z. Wang, Chem. Eng. J., 2021, 421, 129824.
9. H. M. Wilson, S. Rahman A. R, A. E. Parab and N. Jha, Desalination, 2019, 456, 85–96.
10. S. Yu, Y. Wei, D. Fang, W. Li, J. Yu, Y. Tang, J. Bai and J. Wu, Desalination, 2024, 586, 117909.
11. Z. Gong, A. Suwardi and J. Cao, Adv. Funct. Mater., 2025, 2423371.
12. P. Tao, G. Ni, C. Song, W. Shang, J. Wu, J. Zhu, G. Chen and T. Deng, Nat. Energy, 2018, 3, 1031–1041.
13. T. Kumar, B. Zeng, H. Hamza, H. Zhao and X. Zhang, Desalination, 2024, 585, 117686.
14. D. Tong and B. Song, Desalination, 2022, 527, 115586.
15. Y. Wei, W. Li, S. Zhang, J. Yu, Y. Tang, J. Wu and S. Yu, Adv. Funct. Mater., 2024, 34, 2401149.
16. Y. Guo, S. Yu, L. Lv, X. Ge, J. Xu, X. Chao, X. Wang, X. Shi, F. Liang, H. D. Yu and D. Shou, Adv. Mater. Technol., 2025, 10, 2500181.
17. Q. Guo, H. Yi, F. Jia and S. Song, Desalination, 2022, 541, 116028.
18. J. Wu, G. Yin, J. Liu, Z.-Z. Yu and X. Li, Mater. Horiz., 2025, 12, 2878–2898.
19. M. Zhu, R. Zhang, S. Yuan and P. Liu, Desalination, 2025, 601, 118559.
20. H. A. M. Saeed, W. Xu and H. Yang, Carbon Energy, 2024, 6, e540.
21. Z. Jiang, S.-H. Ho, X. Wang, Y. Li and C. Wang, Environ. Pollut., 2021, 290, 118087.
22. Y. Liu, H. Liu, J. Xiong, A. Li, R. Wang, L. Wang, X. Qin and J. Yu, Chem. Eng. J., 2022, 427, 131539.
23. O. Ufuoma Abigail and O. Victoria Ikpemhinoghena, Int. J. Sci. Res. Arch., 2023, 8, 234–243.
24. M. Jin, X. Qu, J. Li, L. Deng, Z. Han, S. Chen and H. Wang, Carbohydr. Polym., 2023, 321, 121324.
25. Y. Liu, S. Yu, R. Feng, A. Bernard, Y. Liu, Y. Zhang, H. Duan, W. Shang, P. Tao, C. Song and T. Deng, Adv. Mater., 2015, 27, 2768–2774.
26. X.-J. Zha, X. Zhao, J.-H. Pu, L.-S. Tang, K. Ke, R.-Y. Bao, L. Bai, Z.-Y. Liu, M.-B. Yang and W. Yang, ACS Appl. Mater. Interfaces, 2019, 11, 36589–36597.
27. Y. Qin, Y. Li, R. Wu, X. Wang, J. Qin, Y. Fu, M. Qin, Z. Wang, Y. Zhang and F. Zhang, Nanomaterials, 2022, 12, 3487.
28. Y. Li, Y. Shi, H. Wang, T. Liu, X. Zheng, S. Gao and J. Lu, Carbon Energy, 2023, 5, e331.
29. O. V. Sedelnikova, D. V. Gorodetskiy, A. G. Kurenya, K. I. Baskakova, E. V. Shlyakhova, A. A. Makarova, G. V. Gorokhov, D. S. Bychanok, P. P. Kuzhir, S. A. Maksimenko, L. G. Bulusheva and A. V. Okotrub, Mater., 2021, 14, 3275.
30. H. Jang, J. Choi, H. Lee and S. Jeon, ACS Appl. Mater. Interfaces, 2020, 12, 30320–30327.
31. Y. Luo, Z. Chen, Q. Li and X. Chen, ACS Appl. Energy Mater., 2021, 4, 9766–9774.
32. H.-r Moon and B. Ryu, Int. J. Precis. Eng. Manuf.-Green Technol., 2024, 11, 1279–1294.
33. L. Meng, D. Cao, J. O. Pedersen, G. Greczynski, V. Rogoz, W. Limbut and M. Eriksson, ACS Appl. Mater. Interfaces, 2025, 17, 39719–39731.
34. X. Li, R. Lin, G. Ni, N. Xu, X. Hu, B. Zhu, G. Lv, J. Li, S. Zhu and J. Zhu, Natl. Sci. Rev., 2018, 5, 70–77.
35. J. Yan, Q. Su, W. Xiao, Z. Wu, L. Chen, L. Tang, N. Zheng, J. Gao and H. Xue, Desalination, 2022, 531, 115686.
36. L. Xiao, W. Xu, L. Huang, J. Liu and G. Yang, Adv. Compos. Hybrid Mater., 2022, 6, 15.
37. C. H. Dreimol, H. Guo, M. Ritter, T. Keplinger, Y. Ding, R. Günther, E. Poloni, I. Burgert and G. Panzarasa, Nat. Commun., 2022, 13, 3680.
38. H. J. Woo, J. Han, S. Ji, B. G. Shin, S. H. Park, S.-G. Lee, C.-W. Lee, E. Hwang, D.-S. Kim, S. Choi, M. S. Jeong, G.-R. Yi, J. Kim and Y. J. Song, ACS Nano, 2024, 18, 12333–12340.
39. T. Lee, J. Kim, I. Koirala, Y. Yang, T. Badloe, J. Jang and J. Rho, ACS Appl. Mater. Interfaces, 2021, 13, 26299–26307.
40. W. Shao, G. Zhu, X. Wang, Z. Zhang, H. Lv, W. Deng, X. Zhang and H. Liang, ACS Appl. Mater. Interfaces, 2023, 15, 932–941.
41. L. Ding, X. Shan, D. Wang, B. Liu, Z. Du, X. Di, C. Chen, M. Maddahfar, L. Zhang, Y. Shi, P. Reece, B. Halkon, I. Aharonovich, X. Xu and F. Wang, Adv. Sci., 2022, 9, 2203354.
42. J. Lin, Z. Peng, Y. Liu, F. Ruiz-Zepeda, R. Ye, E. L. G. Samuel, M. J. Yacaman, B. I. Yakobson and J. M. Tour, Nat. Commun., 2014, 5, 5714.
43. T. Yamashita and P. Hayes, Appl. Surf. Sci., 2008, 254, 2441–2449.
44. D. Wang, X. Wang, Z. Li, M. Chi, Y. Li, Y. Liu and Y. Yin, ACS Nano, 2018, 12, 10949–10956.
45. A. Kaniyoor and S. Ramaprabhu, AIP Adv., 2012, 2, 032183.
46. T. S. D. Le, S. Park, J. An, P. S. Lee and Y. J. Kim, Adv. Funct. Mater., 2019, 29, 1902771.
47. S. Zhu, Z. Lei, Y. Dou, C.-W. Lou, J.-H. Lin and J. Li, Chem. Eng. J., 2023, 452, 139403.
48. A. P. Grosvenor, B. A. Kobe, M. C. Biesinger and N. S. McIntyre, Surf. Interface Anal., 2004, 36, 1564–1574.
49. C. Lin, Z. Pei, J. Lan, H. Huang, X. Ma, S. Cao and C. Wang, Int. J. Biol. Macromol., 2025, 291, 139097.
50. V. Mohan, P. Pazhamalai, K. Krishnamoorthy and S.-J. Kim, Surf. Interfaces, 2024, 51, 104553.
51. J. M. Song, M. Latif, Y. Jiang, Z. Ounaies and J. Kim, Mater. Today Nano, 2024, 26, 100475.
52. K. Ratajczak and M. Stobiecka, Carbohydr. Polym., 2020, 229, 115463.
53. A. Du, Z. Lu and L. Hua, Compr. Rev. Food Sci. Food Saf., 2024, 23, e13421.
54. J. H. S. Almeida, S. D. B. Souza, E. C. Botelho and S. C. Amico, J. Mater. Sci., 2016, 51, 4697–4708.
55. Y. Pang, X. Chu, L. Song, L. Jin, C. Ma, Y. Wu, L. Li, Y. Peng, X. Zheng, F. Wang, S. Wu, Z. Shen and H. Chen, Chem. Eng. J., 2024, 479, 147891.
56. G. W. Ho, Y. Yamauchi, L. Hu, B. Mi, N. Xu, J. Zhu and P. Wang, Nat. Water, 2025, 3, 131–134.
57. H. Shan, Z. Ye, J. Yu, R. Wang and Z. Xu, Device, 2023, 1, 100065.
58. J. Zhou, Y. Gu, P. Liu, P. Wang, L. Miao, J. Liu, A. Wei, X. Mu, J. Li and J. Zhu, Adv. Funct. Mater., 2019, 29, 1903255.
59. M.-J. Jeong, A.-L. Dupont and E. R. de la Rie, Cellulose, 2012, 19, 1135–1147.
60. J. Yao, M. Sun and Y. Yan, Polym. Degrad. Stab., 2024, 229, 110953.

---