Metal‐Free Ternary‐Composite Aerogel Based on Graphitic Carbon Nitride, Graphene Oxide, and β‐Cyclodextrin for Synergistic Adsorption and Photodegradation of Organic Dyes

ABSTRACT

The adsorption‐coupled degradation of dyes offers an attractive strategy for the remediation of industrial dye wastewater. Herein, a metal‐free, porous ternary‐composite aerogel was rationally designed and constructed using graphitic carbon nitride (g‐C₃N₄), graphene oxide (GO), and β‐cyclodextrin (β‐CD). The g‐C₃N₄/GO/β‐CD aerogel (denoted as CNGOCD) exhibited excellent cooperative adsorption enrichment and photo‐driven degradation capabilities for several organic dyes, including rhodamine B, methylene blue, and neutral red, achieving an adsorption capacity of 653.2 mg/g and a removal rate of 97.1%. These results outperformed traditional reduced graphene oxide (rGO) and rGO/β‐CD aerogels. After five cycles, the CNGOCD aerogel maintained stable performance without a significant decrease in either adsorption or photocatalytic activity. Spectral analysis, electrochemical measurements, and density functional theory (DFT) calculations were employed to elucidate the adsorption mechanisms and the enhanced photocatalytic performance of the aerogel. Our findings suggest that GO enhances charge separation in g‐C₃N₄ to boost photocatalytic activity, while β‐CD prevents GO stacking, creating a porous architecture rich in active sites for rapid adsorption. Furthermore, this metal‐free aerogel is eco‐friendly due to no risk of heavy metal leaching, causing secondary pollution. The insights gained from this work provide a solid foundation for designing advanced, multi‐purpose materials aimed at effective pollutant decontamination.

A novel metal‐free g‐C₃N₄/GO/β‐CD aerogel exhibits a high adsorption capacity of 653.2 mg/g for RhB. The synergistic adsorption‐degradation effect of the aerogel enables the complete elimination of organic dyes. Fixed‐bed and recyclability tests demonstrate its potential for continuous industrial wastewater treatment. DFT calculations and experimental analysis reveal the underlying adsorption‐degradation mechanism.

1. Introduction

Due to its high chromaticity, elevated chemical oxygen demand (COD), and low biodegradability, industrial dye wastewater represents a major environmental and health concern [1]. In the aquatic environment, its high chromaticity greatly hinders the photosynthesis of aquatic plants. Meanwhile, the high COD rapidly depletes dissolved oxygen in water, disrupting the aquatic ecosystem. When dye wastewater is discharged untreated, it causes harmful substances to accumulate, threatening human health. Currently, there are various treatment methods for industrial dye wastewater, such as physicochemical, biological, electrochemical, and advanced oxidation techniques [2, 3]. Among them, the common adsorption method uses adsorbents to capture dye molecules, which features simple operation, low cost, good treatment effect, and recyclability [4]. However, the adsorption method has limitations like limited adsorption capacity and potential for secondary pollution. Conversely, integration adsorption and degradation process, the adsorption‐degradation method not only attains good treatment results and avoids secondary pollution, but also extends the lifespan of the adsorbent [5].

Integrating diverse water treatment methods leverages their unique strengths and can overcome the limitations of single‐method approaches, such as low efficiency, high costs, and poor sustainability. Notably, the combination of adsorption and photocatalysis has emerged as a promising strategy for cost‐effective and eco‐friendly contaminant removal [6, 7, 8]. Reactive oxygen species (ROS), which are crucial for photocatalytic degradation, have an extremely short lifespan [9]. One of the decisive factors in the success of photodegradation hinges on the adsorption rate at which target pollutants attach to the surface of the photocatalyst. Currently, metal‐oxide photocatalysts such as titanium dioxide (TiO₂), bismuth oxyiodide (BiOI), and cadmium sulfide (CdS), despite boasting excellent photocatalytic capabilities, exhibit subpar adsorption properties [10, 11, 12, 13]. This drawback impedes the efficient utilization of ROS for pollutant degradation and diminishes photocatalytic efficiency. In addition, metal‐oxide photocatalysts may cause secondary contamination through heavy metal leaching. Photocatalysts with good adsorption properties can quickly concentrate target contaminants on or near the catalyst surface, thus achieving enrichment and in situ degradation [6]. The porosity and high specific surface area of adsorption‐type photocatalytic materials facilitate the adsorption of organic dyes, enhancing photocatalytic degradation by trapping more visible light and facilitating the separation of electron‐hole pairs. This integrated approach allows for the simultaneous resolution of pollutant enrichment and separation in the water environment all in a single step. Among these materials, graphitic carbon nitride (g‐C₃N₄) is a well‐known visible light responsive photocatalyst due to its high stability, adjustable bandgap, and environmental friendliness [14]. However, the practical application of pristine g‐C₃N₄ is often hindered by several intrinsic shortcomings, principally the rapid recombination of photogenerated electron‐hole pairs, a limited specific surface area, and inadequate light harvesting capability. According to previous reports [7, 15], the above shortcomings can be overcome by constructing three‐dimensional (3D) porous structures to obtain a large accessible surface area, expose more reactive sites, and open interconnected structures to form multi‐light reflections to improve light capture capability.

The creation of superior 3D porous matrices demands a precise combination of novel material systems and robust synthesis protocols. Graphene oxide (GO) nanosheets can be easily self‐assembled into graphene aerogels through π–π stacking and hydrogen bonding [16]. However, when pure GO is used to assemble 3D aerogels, it is difficult to achieve high stability, high porosity, and abundant active sites simultaneously due to the substantial π–π stacking among GO nanosheets during the assembly phase. Many attempts have been made to solve the problem of severe π–π stacking in 3D aerogel formation [17, 18, 19]. A common approach was adding chemical or physical additives like DNA, metal cations, or polymeric chains, aiming to inhibit this stacking [20, 21, 22]. However, these dopants often have disadvantages that are often expensive, toxic, or unstable, raising costs and environmental risks. Besides, they mainly functioned as stabilizers without adding extra value. As a result, the modified 3D aerogels still did not have great adsorption and photocatalysis ability. Consequently, there is an urgent need for an efficient and mild preparation method that can endow aerogels with high stability, strong adsorption capacity, and excellent photocatalytic performance simultaneously. More recent strategies [23, 24, 25], including the use of sustainable biomass to construct photoactive aerogels, hybridization to enhance electronic properties and dye uptake, and meticulous pore engineering for improved mass transfer, have advanced the field. However, achieving a fully metal‐free system that simultaneously integrates high‐capacity adsorption, rapid photocatalysis, and eco‐friendly safety remains a challenge.

Due to their abundant hydroxyl groups and unique cyclic structure, natural oligosaccharides known as cyclodextrins (α‐, β‐, γ‐CD) have been successfully adapted into functional materials for various applications [26, 27, 28, 29, 30, 31]. Among them, β‐CD is the most cost‐competitive and is widely adopted for assembling porous adsorption systems in combination with GO and other agents [6, 32, 33, 34, 35, 36]. Herein, we designed and fabricated a series of metal‐free g‐C₃N₄/GO/β‐CD porous ternary‐composite aerogels by a facile, eco‐friendly method (Scheme 1). In contrast to many existing carbonaceous aerogels, our CNGOCD system leverages the unique host‐guest inclusion capability of β‐CD to achieve efficient synergy between adsorption enrichment and in situ photodegradation. On the one hand, the high electron mobility of GO facilitates charge transfer, thereby hindering the rapid recombination of photogenerated electrons and holes g‐C₃N₄. On the other hand, the incorporation of β‐CD can disrupt the stacking of GO and GO, GO and g‐C₃N₄, improve the composite material's stability, and increase the active sites to improve the adsorption capacity. The formation of the 3D composite can enhance the light reflection ability, which in turn boosts the photocatalytic activity. Moreover, the improvement in adsorption capacity can accelerate the rate of photocatalytic surface reactions. Through systematic experiments and theoretical simulations, we elucidated the structure‐property relationships governing the adsorption‐degradation synergy and charge transfer mechanisms. The developed aerogel achieved excellent removal efficiency for multiple dyes (rhodamine B‐RhB, methylene blue‐MB, and neutral red‐NR) while maintaining stable recyclability, offering new insights into the design of metal‐free multifunctional materials for environmental remediation. This work not only advances the fundamental understanding of adsorption‐coupled photocatalysis but also provides a sustainable strategy for addressing complex wastewater treatment challenges.

SCHEME 1.

Schematic illustration of the synthetic procedure of CNGOCD aerogels via hydrothermal self‐assembly and the synergistic adsorption‐photodegradation functionality of the 3D porous network.

2. Materials and Methods

2.1. Materials

All reagents were of analytical purity and utilized as received without second purification. Natural graphite flake (800 mesh) were sourced from Shenzhen Hanhui Graphite Co., Ltd. Reagents including concentrated sulfuric acid (H₂SO₄, 98 wt.%), hydrochloric acid (HCl), sodium carbonate (Na₂CO₃), sodium nitrate (NaNO₃), sodium sulfate (Na₂SO₄), potassium permanganate (KMnO₄), and hydrogen peroxide aqueous solution (H₂O₂, 30 wt.%) were procured from Yonghua Chemical Co., Ltd. Additionally, Melamine (C₃H₆N₆, 99%), β‐cyclodextrin (β‐CD, 98%), L‐ascorbic acid (L‐AA, 99.5%), sodium dihydrogen phosphate (NaH₂PO₄, 99%) were acquired from Aladdin (Shanghai, China). Deionized (DI) water was used for all preparations.

2.2. Synthesis of GO, g‐C₃N₄, and CNGOCD Composites Aerogels

2.2.1. Synthesis of GO and g‐C₃N₄

The preparation of GO was conducted in accordance with the modified Hummers’ method [37]. For the synthesis of g‐C₃N₄ nanosheets, a facile thermal oxidation etching strategy was adopted [38]. First, bulk g‐C₃N₄ was synthesized by heating melamine in static air at a temperature of 500°C for 3 h with a ramp rate of 5°C min^{− 1}. This yellow intermediate product was collected, ground, and further exfoliated into nanosheets by heating the bulk powder at 550°C for 3.5 h under identical ramp conditions.

2.2.2. Synthesis of CNGOCD Composite Aerogels

CNGO aerogel was prepared via a single‐step hydrothermal method, based on the method detailed in a prior study [7]. Initially, 10 mL of a 16 mg/mL g‐C₃N₄ nanosheet suspension was mixed with 10 mL of a 4 mg/mL GO nanosheet suspension. The mixture underwent 60‐min sonication. Then, the resulting gray‐stable suspension was transferred to a 50 mL Teflon vessel and heated (180°C, 12 h). After being washed three times with water, the black column‐shaped product was freeze‐dried in a vacuum freeze‐dryer for 48 h, yielding the 4CN1GO0CD aerogel. For the preparation of CNGOCD ternary composite aerogel, 40 mL of original CN‐GO suspension (CN 4.0 g/L, GO 1.0 g/L) was mixed with the β‐CD (0.5 g/L), NaH₂PO₄ (0.11 g/L), and L‐AA (10 g/L). The mixture underwent 60‐min sonication. Then, the resulting suspension was transferred to a 50 mL Teflon vessel and heated at 90°C for 4 h. After being washed three times with water/ethanol (1:9, v/v), the black column‐shaped product was freeze‐dried in a vacuum freeze‐dryer for 48 h, yielding the 4CN1GO0.5CD aerogel. Except for changing the concentration of GO (1.0, 2.0, or 3.0 g/L) and β‐CD (0.5, 1.0, or 2.0 g/L), the synthesis methods of 4CN1GO, 4CN2GO, or 4CN3GO series aerogels are the same as those for the 4CN1GO0.5CD aerogel. Unless otherwise specified, the numbers all represent the final concentrations of substances in the system.

2.3. Material Characterization

Sample characterization utilized the following instrumentation: SEM (FEI Quanta 250) for morphology; FT‐IR (Thermo Fisher Scientific Nicolet iS20) and Raman (Horiba LabRAM HR Evolution, He‐Ne laser) for functional group and structural analysis; and XPS (Shimadzu/Kratos Axis Supra, Mg Kα, 1253.6 eV) for surface chemistry. XRD patterns were recorded using a Rigaku D/max‐2200PC diffractometer (Cu Kα, λ = 0.154 nm, 2θ = 5°–60°, step: 0.02°). The specific surface area and pore size distribution were determined by N₂ adsorption–desorption at 77 K using a Micromeritics ASAP 2460. UV–vis DRS data were obtained via a Shimadzu UV‐3600i Plus spectrophotometer with a BaSO₄ reference. Electrochemical studies (Mott‐Schottky plots) were performed on a CHI660E workstation in 0.5 M Na₂SO₄ solution using a conventional three‐electrode cell (Pt counter, Ag/AgCl reference, FTO working electrode).

2.4. Batch Adsorption Experiments

Stock solutions (3000 mg/L) of three representative industrial dye contaminants, RhB, MB, and NR, were prepared and subsequently diluted to the required working concentrations. To isolate the adsorption mechanism and eliminate interference from potential photocatalytic degradation, all batch experiments were conducted under light‐shielded conditions. The equilibrium adsorption capacity of the aerogel was quantified based on the variation in pollutant concentration before and after treatment, calculated via Equation (1):

$$Q_e=\left(C_0-C_e\right)\times V\times m^{-1}$$ (1)

here, Q_e (mg/g) denotes the saturated adsorption capacity of aerogel at equilibrium; C₀ and C_e (mg/L) refer to the dye concentrations at the initial stage and at equilibrium, respectively; V (L) refers to the volume of pollutant solution, and m (g) is the dosage of aerogel.

2.4.1. Adsorption Kinetic Experiments

To elucidate the adsorption rate and mechanism, batch adsorption experiments were conducted by varying the pH values and contact times. In the investigation of the pH effect on adsorption performance, 10 mg of aerogel was added to a solution with an initial dye concentration—RhB (150 mg/L), MB (100 mg/L), and NR (100 mg/L). The pH of the solution was adjusted from 3 to 10, and the experiment was conducted at 298.15 K. Adsorption kinetics were evaluated by exposing 10 mg of adsorbent to 10 mL of mixed pollutants (cumulative concentration: 150 mg/L) for varying intervals (10 s—4 h). For time points ≤ 60 s, an aliquot of the suspension was rapidly withdrawn with a syringe and immediately filtered through a 0.22 µm polytetrafluoroethylene syringe filter. The filtrate was directly collected for absorbance measurement. To elucidate the adsorption mechanism, the obtained kinetic data were simulated using the classical pseudo‐first‐order and pseudo‐second‐order equations, represented by Equations (2) and (3).

Pseudo‐first‐order:

(2)

Pseudo‐second‐order:

$$t/Q_t=1/\left(K_2{Q}_e^2\right)+t/Q_e$$ (3)

here, Q_t and Q_e (mg/g) represent the adsorption capacities at a specific time t and at equilibrium, respectively. K₁ (min⁻¹) and K₂ (g mg⁻¹ min⁻¹)denote the rate constants corresponding to the pseudo‐first‐order and pseudo‐second‐order kinetic models.

2.4.2. Adsorption Isotherm Experiments

To assess the saturation limit of the aerogels toward the target dyes (RhB, MB, and NR), a detailed isotherm study was performed. The experimental setup involved treating varying concentrations of dye solutions (50–1000 mg/L, 10 mL) with a fixed dosage of adsorbent (10 mg). After equilibrium was established and phases were separated, the remaining dye concentrations were detected via a multifunctional microplate reader. The interaction mechanisms were subsequently investigated by fitting the data to the classical Langmuir and Freundlich isotherm equations, provided in Equations (4) and (5):

Langmuir model:

$$Q_e=\left(Q_m\times K_L\times C_e\right)/\left(1+K_L\times C_e\right)$$ (4)

Freundlich model:

$$Q_e=K_F\times C_e^{-n}$$ (5)

here, C_e (mg/L) and Q_e (mg/g) represent the pollutant concentration and the corresponding experimental capacity of aerogel at equilibrium, respectively; Q_m (mg/g) refers to the theoretical maximum adsorption capacity calculated via the Langmuir model, while K_L and K_F are the characteristic constants derived from the Langmuir and Freundlich fitting processes, respectively, with n⁻¹ reflecting the heterogeneity or intensity of the adsorption process.

2.4.3. Fixed‐Bed Column Experiments

The fixed‐bed experiments were carried out using a manually prepared column (diameter: 25 mm, length: 100 mm). A cotton ball was placed at the bottom and top of the column respectively. The column was filled with CNGOCD aerogel to a height of 50 mm. RhB solution (50 mL, 5 mg/L) was continuously pumped into the column at a flow rate of 2 mL/min. The initial and final absorbance values of the RhB solution were compared.

2.5. Photocatalytic Activity

Photocatalytic activity of the synthesized CNGOCD aerogel was investigated by monitoring the decomposition of RhB (300 mg/L), MB (50 mg/L), and NR (25 mg/L) under 45 W LED white light (λ > 400 nm). In a typical assay, the photocatalyst (1.0 g/L) was suspended in a 50 mL organic dye solution within a 100 mL quartz reactor. Prior to light exposure, the system was magnetically stirred in the dark for 60 min to saturate the photocatalyst surface with dye molecules. During the reaction, the degradation efficiency was tracked by sampling 2.0 mL of the mixture and measuring the absorbance using a multifunctional microplate reader at the maximal absorption wavelengths of RhB (553 nm), MB (664 nm), and NR (540 nm). The total organic carbon (TOC) concentration during degradation was measured at 1, 2, 4, and 6 h using a TOC‐L CPH analyzer after filtration through 0.22 µm membrane. All the samples were tested in parallel three times, and data are reported as mean ± standard‐deviation (SD) (n = 3).

2.6. Theoretical Computations

To unravel the mechanism governing aerogel assembly and its adsorption‐photocatalysis mechanism for RhB, MB, and NR, DFT studies were performed. Using the B3LYP/6‐31G* basis set within Gaussian 16W [39], the system underwent geometry optimization and frequency analysis. Natural Bond Orbital (NBO) analysis [40] was subsequently employed to map electron spin densities and charge populations across GO, g‐C₃N₄, β‐CD, and the target dyes. Furthermore, frontier molecular orbitals (HOMO and LUMO) were visualized utilizing Multiwfn [41, 42] and VMD [43] software.

A critical component of this study involved predicting reactive sites using Fukui functions, which assess susceptibility to nucleophilic ($f_K^{+}$), electrophilic ($f_K^{-}$), and radical ($f_K^0$) attacks. These are calculated as follows:

Nucleophilic attack:

$$f_K^{+}=q_N^K--q_{N+1}^K$$ (6)

Electrophilic attack:

$$f_K^{-}=q_{N-1}^K--q_N^K$$ (7)

Radical attack:

$$f_K^0=\left(q_{N-1}^K--q_{N+1}^K\right)/2$$ (8)

In these equations, q^K denotes the atom charge of atom K in the specified state. Generally, a higher Fukui index correlates with increased reactivity. By analyzing the Fukui index, the vulnerable regions of substrate and organic dye molecules were identified, thereby elucidating the underlying mechanism of aerogel synthesis and dye photo‐degradation.

3. Results and Discussion

3.1. Structural Characterizations and Chemical Composition Analysis

Figure 1a illustrates the morphological characteristics of freeze‐dried aerogels following hydrothermal synthesis. The CNGOCD series aerogels exhibited successful polymerization, retaining an intact cylindrical morphology. To analyze crystallographic features, XRD was recorded for CNGOCD composites alongside pure g‐C₃N₄ and GO (Figure 1b and Figure S1). Pristine g‐C₃N₄ displayed two prominent reflections at 13.1° and 27.8°, assigned to the (100) and (002) crystallographic planes characteristic of graphitic materials [7]. The intense peak at 27.8° corresponds to interlayer stacking in conjugated aromatic networks. Pure GO exhibited a singular sharp diffraction feature at 10.8°, attributed to its (002) plane [44]. In composite samples, the persistence of the 27.8° peak confirmed structural retention of g‐C₃N₄. Notably, increasing GO incorporation progressively attenuated the 10.8° diffraction signal in CNGOCD aerogels, suggesting partial reduction of GO to reduced GO (rGO) during hydrothermal treatment and subsequent graphitic reorganization. Additional minor peaks emerged at 2θ values of 19.7° and 21.6°, potentially corresponding to β‐CD crystalline features (Figure S1c).

FIGURE 1.

Structural and chemical characterization of the CNGOCD aerogels. Image of freeze‐dried GNCOCD aerogel (a); XRD patterns of GNCOCD aerogel with varying GO and β‐CD contents (b); XPS patterns of 4CN1GO2CD aerogel (c); High‐resolution XPS spectra of (d) C 1s, (e) N 1s, and (f) O 1s for 4CN1GO2CD.

XPS analysis was conducted to elucidate elemental states and components in CNGOCD composites (Figure 1c–f, Figures S2, S3, and Table S1). Wide‐scan spectra confirmed the existence of carbon, nitrogen, and oxygen constituents. High‐resolution C1s spectra for 4CN1GO2CD (Figure 1d) resolved into three distinct components through Gaussian deconvolution: sp²‐hybridized carbon bonds (C─C/C═C, 284.4 eV), oxygen‐containing functional groups (C─O, 286.7 eV), and aromatic carbons coordinated with nitrogen (N─C═N, 288.3 eV). The latter two features align with g‐C₃N₄ signatures, while the C─O component originates from residual oxygen moieties in partially reduced GO, corroborated by the O1s peak at 529.0 eV (Figure 1f). Nitrogen speciation analysis revealed comparable spectral profiles between composites and pure g‐C₃N₄, with N 1s spectra deconvoluting into three states: aromatic N─C bonding (398.7 eV), tertiary nitrogen N─(C)₃ (399.7 eV), and N─H groups (400.0 eV) [38]. These observations, combined with the identified sp²‐hybridized carbon‐nitrogen configurations, substantiate the preservation of heptazine (C₆N₇) units—fundamental structural motifs of g‐C₃N₄. In Table S1, elemental composition analysis indicated that increasing the initial concentration of GO led to a significant reduction in the oxygen mass fraction, accompanied by a marked increase in the nitrogen mass fraction. This observation can be attributed to: (1) the enhanced reduction of GO at higher concentrations promoted a more thorough elimination of oxygen‐containing functional groups, thereby decreasing the oxygen content; (2) the increased availability of GO facilitated its synergistic interaction with g‐C₃N₄, which promoted the formation of stable covalent/non‐covalent bonding configurations. These bonding interactions not only stabilized the hybrid structure but also retained more nitrogen species derived from the g‐C₃N₄ matrix, resulting in the observed increase in nitrogen content.

Figure S1d and S4a show the FT‐IR spectroscopy of pure g‐C₃N₄, GO, and CNGOCD. For pure g‐C₃N₄, its FT‐IR spectrum displayed three characteristic absorption bands. The band above 3000 cm⁻¹ was ascribed to N─H stretching [45]. A series of peaks between 1200 and 1650 cm⁻¹ corresponded to the stretching modes of CN heterocycles, and a sharp peak below 1000 cm⁻¹ was attributed to the breathing mode of s‐triazine units [7, 38]. Regarding GO spectrum, three oxygenated functional peaks were detected at 1733, 1624, and 1074 cm⁻¹. These peaks respectively represented the stretching vibration of C═O, epoxy C─O, and alkoxy C─O [46]. Notably, all of these characteristic peaks were also present in the spectrum of CNGOCD samples, confirming the successful integration of these components. Moreover, in the CNGOCD spectrum, the incorporation of β‐CD is evidenced by the emergence of new signals at 2960 cm^{− 1} (aliphatic C─H stretching), 1170 cm^{− 1} (symmetric C─O─C stretching), and the bands at 969 and 1050 cm^{− 1} (ring vibrations and C─O/C─C stretching) [47]. To evaluate the structural evolution and electronic properties, Raman analysis was conducted (Figure S4b). The spectra exhibit the standard G peak (1580 cm^{− 1}, sp² carbon vibration) and D peak (1350 cm^{− 1}, defect‐induced scattering) [48]. The ID/IG ratio, a standard metric for quantifying the degree of disorder or functionalization, was calculated for the varying samples. The values were found to increase from 1.07 for 4CN1GO0CD to 1.20 for 4CN3GO2CD. This notable rise in the ID/IG ratio following hydrothermal polymerization indicates a significant restoration of the conjugated network and a reduction degree superior to that of the original GO.

The morphologies of GO, g‐C₃N₄, and CNGOCD series aerogels were analyzed via SEM (Figure 2 and Figure S5). Figure 2a–c shows the SEM images of the freeze‐dried aerogels after the hydrothermal preparation. In contrast to pure GO, g‐C₃N₄, CNGA, and β‐CD‐CA‐rGO aerogels reported in literatures [7, 32, 33], The CNGOCD aerogel exhibited a well‐defined porous network, whereas the CNGA and β‐CD‐CA‐rGO counterparts showed compact and irregular morphologies. The lack of porosity in the latter samples suggests a significantly lower polymerization efficiency, hindering the formation of a stable 3D framework. Moreover, from the SEM images, as the concentrations of GO and CD increased, the interlayer structure of the aerogel samples became denser, and the pore size became smaller. The porous structure of CNGOCD aerogels was characterized by nitrogen adsorption–desorption measurements (Figure S6 and Table S2. All samples exhibit Type IV isotherms with H3‐type hysteresis loops, confirming their mesoporous nature. The BET specific surface areas range from 160.61 to 195.91 m²/g, with total pore volumes of 0.55–0.68 cm³/g and average pore diameters of 12.85–15.07 nm. The BJH pore size distribution shows a dominant peak at 30–40 nm. With increasing GO content, the specific surface area increases slightly while the average pore diameter decreases, consistent with SEM observations. This well‐developed porous network provides abundant active sites and facilitates mass transfer for efficient dye removal.

FIGURE 2.

SEM images of CNGOCD aerogels with systematically varied compositions, revealing the evolution of the 3D porous network. (a–c) 4CN1GO series aerogels with increasing β‐CD content (0.5CD, 1CD, 2CD); (d–f) 4CN2GO series aerogels; (g–i) 4CN3GO series aerogels.

3.2. Adsorption Performance

3.2.1. Adsorption Kinetics and Adsorption Isotherm of CNGOCD Aerogels

The adsorption capacities of the synthesized CNGOCD aerogels were assessed using RhB as the model dye to investigate the adsorption kinetics and isotherm models (Figure 3). All the aerogels rapidly adsorbed RhB within the initial 10 min (Figure 3a–c). The extremely fast removal rates of CNGOCD aerogels were mainly attributed to hydrogen bonding, hydrophobic and electrostatic interactions, which were induced by the plentiful ─OH and ─COOH active groups. Beyond the initial hour, the uptake rate slowed significantly, with the adsorption capacity reaching a plateau and maintaining stability up to 240 min. To elucidate the underlying mechanism, the experimental time‐dependent data were simulated using both pseudo‐first‐order and pseudo‐second‐order kinetic equations (parameters listed in Table S3). Notably, the pseudo‐second‐order model yielded a superior determination coefficient (R²), providing strong statistical evidence that the adsorption process on CNGOCD aerogels is primarily governed by chemisorption [34].

FIGURE 3.

Adsorption kinetics and isotherms of CNGOCD aerogels for RhB. The pseudo‐first‐order and pseudo‐second‐order kinetic model fitting of CNGOCD, 4CN1GO in (a), 4CN2GO in (b), and 4CN3GO in (c), respectively. Langmuir and Freundlich isotherm fittings for the corresponding aerogel series (d–i).

Two well‐known adsorption isotherms, the Langmuir and Freundlich models, were also applied to elucidate the adsorption behavior of CNGOCD aerogels (Figure 3d–i and Figure S8a–d). According to the Langmuir model fitting results, the maximum adsorption capacity of CNGOCD aerogels for RhB could reach as high as 653.2 mg/g. Simultaneously, a comparative analysis of the regression data (Table S4) reveals that the Freundlich model yields consistently higher correlation coefficients (R²) than the Langmuir model fitting. This statistical superiority indicates that the adsorption of dyes onto the CNGOCD aerogel is not limited to a monolayer coverage but rather follows a multilayer sorption process on a heterogeneous surface [35]. Physically, this implies that the adsorption sites on the aerogel are energetically non‐equivalent, likely due to the complex composite structure. Furthermore, the CNGOCD aerogel's excellent hydrophilicity and the abundance of oxygen‐containing functional groups significantly reduced mass transfer resistance, thereby facilitating the diffusion of dye molecules into the porous network. Mechanistically, the high adsorption capacity is attributed to a multi‐force synergy, beyond the primary electrostatic attraction between the negatively charged aerogel surface and cationic dyes, the process is reinforced by hydrogen bonding (between ─OH/─C═O and dye molecules) and the unique host‐guest inclusion effects provided by the β‐CD cavities. Overall, the adsorption performance of the 4CN1GO series aerogels was superior to that of the other two aerogels. As observed from the SEM images, the 4CN1GO series aerogels had a larger and more regular pore structure, which might account for its better adsorption performance. Remarkably, the CNGOCD aerogels achieved an exceptional maximum adsorption capacity of 653.2 mg/g, substantially outperforming many absorbent composites in previously literatures (Table S5). Furthermore, the adsorption properties of four aerogels for MB and NR were evaluated (Figure S7a,b and Tables S6–S9). CNGOCD aerogels also exhibited good adsorption capacity for MB and NR. The adsorption kinetics experiments and adsorption isotherm simulation revealed that CNGOCD aerogels adsorbed MB through multilayer chemisorption, while for NR, single‐layer chemisorption was predominant.

3.2.2. pH‐Dependent Adsorption Behavior of CNGOCD Aerogels

To assess the stability of the CNGOCD aerogels, the aerogels demonstrating optimal removal efficiency for the three dyes (4CN2GO1CD for RhB, 4CN1GO2CD for MB, and 4CN1GO2CD for NR) were selected for adsorption testing across varying pH levels (Figure S7c). RhB, MB and NR are cationic below a pKa value of 3.7, 1.3, and 7.4, respectively. As shown in Figure S7c, the CNGOCD aerogels maintained a superior removal efficiency (> 90%) for both RhB and NR across a broad pH spectrum, exhibiting negligible sensitivity to acidity or alkalinity. This pH‐independence suggests that for RhB and NR, the adsorption mechanism is not solely governed by electrostatic forces; instead, robust non‐electrostatic interactions–such as hydrogen bonding, π–π stacking, and host‐guest inclusion by β‐CD–likely act as the dominant driving forces, ensuring stability even when surface charges fluctuate. Conversely, the adsorption of MB displayed a strong positive correlation with pH, improving progressively from acidic to basic conditions. This trend indicates that electrostatic attraction plays a pivotal role in capturing MB, as the deprotonation of the aerogel surface at higher pH levels enhances its affinity for the cationic MB molecules.

3.3. Assessment of Photocatalytic Activity

3.3.1. Study of Photocatalysis Performance Under Visible Light

A two‐stage experimental protocol was employed to assess the synthesized CNGOCD aerogel's photocatalytic performance under visible light irradiation. To clarify the respective contributions of physical adsorption from chemical photodegradation in removing RhB as a model pollutant, adsorption equilibrium experiments for CNGOCD aerogels were first conducted in the dark before initiating photocatalysis. As previously mentioned, CNGOCD aerogels achieved fast RhB removal within 60 min. Subsequently, the system was exposed to visible light irradiation for 300 min to trigger the photocatalytic reaction. The results showed that all CNGOCD aerogels exhibited relatively high adsorption efficiencies (> 60%) for RhB (Figure 4b). Under natural light conditions, the CNGOCD aerogel demonstrated a synergistic adsorption‐degradation removal efficiency of RhB reaching up to 97% (Figure 4c). Additionally, most ternary composite aerogels displayed superior RhB adsorption and degradation capabilities under visible light compared to the binary composite aerogel (4CN1GO0CD). This enhancement is attributed to the dual‐functional role of β‐CD. First, β‐CD acts as a molecular linker that optimizes the interfacial contact between GO and g‐C₃N₄, thereby facilitating more efficient transfer of charge carrier and suppressing the recombination of electron‐hole. Second, the hydrophobic cavities of CD serve as molecular traps via host‐guest interactions, effectively concentrating dye molecules near the catalytic active sites.

FIGURE 4.

Photocatalytic degradation performance of CNGOCD aerogels toward RhB under visible light irradiation. Real photo of aerogel adsorption‐degradation of RhB (a) and the dynamic degradation and removal ratio of RhB under visible light (b, c). The fittings with the pseudo‐first‐order kinetic model are shown in (d–f) for 4CN1GOxCD, 4CN2GOxCD, and 4CN3GOxCD series of aerogels, respectively. Data are presented as mean ± SD (n = 3). Dark adsorption was carried out for 60 min prior to light exposure to reach adsorption–desorption equilibrium.

The reaction kinetics were quantitatively resolved using the pseudo‐first‐order model, as detailed in Figure 4d–f and Table S10. The k₁ parameter is directly proportional to the degradation rate. In comparison with the binary composite aerogel 4CN1GO0CD, the majority of ternary composite aerogels exhibited higher k₁ values for RhB degradation. Above results indicated that ternary composites possessed stronger dye removal capacities than the binary counterpart. Notably, the 4CN3GO series aerogels generally showed relatively higher k₁ values, implying their enhanced photocatalytic degradation capabilities. These results are consistent with the higher reduction degrees revealed by XPS and Raman spectroscopy characterizations, further validating the structural‐property relationships in these materials.

Subsequently, four aerogels were selected to investigate their adsorption‐degradation performance toward MB and NR (Figure 5a–e, Tables S11 and S12). Experimental results demonstrated that 4CN1GO2CD exhibited the optimal adsorption‐degradation efficiency for both dyes, with removal rates exceeding 98%. Additionally, all ternary composite aerogels outperformed 4CN1GO0CD in dye removal, indicating that ternary composites possess effective elimination capabilities toward other dyes such as MB and NR, thus highlighting their general applicability in dye degradation. The adsorption‐degradation capacity of aerogel toward mixed dyes was also evaluated by treating a mixture of RhB, MB, and NR (Figure 5f). During the degradation process under visible light, after 480 min of irradiation, the removal efficiency of MB exceeded 70%, while the combined removal efficiency for RhB and NR exceeded 90%. The removal performance in the multi‐component system was inferior to that observed in single‐solute solutions. This decline is primarily attributed to the formation of an electrostatic barrier. As dye molecules accumulate on the aerogel surface, they generate repulsive charges that inhibit the approach and adsorption of subsequent pollutant molecules. As shown in Table S13, when benchmarked against g‐C₃N₄‐based photocatalysts, the CNGOCD ternary composite demonstrates superior visible‐light photocatalytic efficiency and enhanced efficacy in dye wastewater remediation. Furthermore, the comparative performance of the CNGOCD aerogel against standard GO and various biomass‐derived aerogels is summarized in Table S14, demonstrating the distinct advantages of our synergistic system. This highlights its competitive edge in harnessing visible light energy and degrading organic pollutants, thereby establishing it as a highly viable candidate for real world environmental remediation.

FIGURE 5.

The photodegradation properties of CNGOCD aerogels toward MB and NR. The dynamic degradation under visible radiation and the linear fitting results based on pseudo‐first‐order kinetic models were presented in (a, b), (c, d) for MB and NR, respectively. The removal ratio of MB and NR and photodegradation property of 4CN3GO2CD aerogel on dye mixture (RhB + MB + NR) were exhibited in (e, f). Data are presented as mean ± SD (n = 3).

3.3.2. Ionic Interference Photodegradation Property of CNGOCD Aerogels

In practical wastewater treatment, the co‐existence of inorganic anions is inevitable and is typically reported to suppress photocatalytic efficiency by competing for active surface sites [49]. To evaluate the robustness of the CNGOCD aerogel against such interferences, a series of simulated degradation assays was conducted in the presence of ubiquitous anions, including Cl⁻, CO₃ ²⁻, NO₃ ⁻, and SO₄ ²⁻ (Figure 6). The degradation efficiency of RhB exhibited negligible fluctuations upon the addition of these ions, suggesting that the CNGOCD aerogel maintains high catalytic activity even in complex ionic matrices, likely due to its strong specific affinity for the RhB molecules, which overcomes anionic competition. However, SO₄ ²⁻, NO₃ ⁻, and Cl⁻ exerted significant adverse effects on the catalytic degradation of MB and NR. Specifically, the presence of SO₄ ²⁻ reduced the catalytic degradation efficiencies of MB and NR to 75% and 60%, respectively, which may be attributed to the adverse impact of the common ion effect on the degradation process. Although most anions tested in the experiment caused some negative effects on the photocatalytic performance of the CNGOCD aerogels, the concentration of anions in real water environments is only 1000 ppm, which is far lower than the conditions set in this study, indicating that the ternary composite aerogel system still exhibits good anti‐interference capability toward these anions.

FIGURE 6.

Ionic interference photodegradation property of CNGOCD aerogels. Influence of coexisting inorganic anions in solutions on the removal efficiency of CNGOCD aerogels toward RhB (a), MB (b), and NR (c). Inorganic salt (NaCl, Na₂CO₃, NaNO₃, and Na₂SO₄) concentration: 0.1 m. Data are presented as mean ± SD (n = 3).

3.3.3. Fixed‐Bed Experiment and Photodegradation Recyclability of CNGOCD Aerogels

For commercial implementation, efficient adsorption capacity in continuous wastewater streams is critical. To evaluate this, a fixed‐bed column system was employed to conduct continuous RhB adsorption experiments. CNGOCD aerogels were packed into a custom‐built column (Figure 7a). In Figure 7b, a RhB influent (5 mg/L) was percolated through the aerogel‐packed column at a steady flow rate (2 mL/min), with the effluent collected after passing through the setup. Spectral analysis of the outlet stream revealed a nearly 99% reduction in RhB absorbance, indicating effective dye removal. Furthermore, the absorption baseline of the treated water remained flat, and no visual turbidity or darkening was observed, suggesting that the CNGOCD aerogel possesses robust mechanical stability, effectively preventing particle leaching or secondary contamination during the water flux (Figure 7c). From an economic perspective, adsorbent reusability is a key criterion for practical utility. The recyclability of CNGOCD aerogels was assessed after elution of RhB degradation by‐products with an ethanol/water (9:1) mixture (Figure 7d). After five regeneration cycles, the aerogels retained high adsorption‐degradation performance, demonstrating long‐term operational durability. These results highlight the potential of CNGOCD aerogels for repeated use in industrial applications, supporting their feasibility for commercial‐scale wastewater treatment systems.

FIGURE 7.

Continuous fixed‐bed adsorption and cyclic reusability of CNGOCD aerogels. (a) Photographic image of CNGOCD aerogel column. (b) Photographic image of the setup for dynamic adsorption using a CNGOCD aerogel‐packed bed column (RhB: 5 mg/L). (c) Spectral comparison of the RhB solution before and after column treatment. (d) Reusability of CNGOCD aerogel with 5 adsorption‐degradation cycles for eliminating RhB. Error bars represent SD from three replicate measurements (n = 3).

3.4. Photodegradation Mechanism

3.4.1. Evaluation of Light Response Performance

To delineate the electronic band structure, specifically the valence band (VB) and conduction band (CB), the flat‐band potential (FB) was determined via Mott‐Schottky plotting. A positive slope observed in CNGOCD aerogels signifies their n‐type semiconducting behavior [50]. FB values for 4CN1GO series aerogels, 4CN2GO2CD, and 4CN3GO2CD are presented in Figure 8a–f. UV–vis Diffuse reflectance spectroscopy (UV–vis DRS) was utilized to characterize the light‐response properties of CNGOCD aerogels (Figure 8g and Figure S9). As shown in Figure S9, for pristine g‐C₃N₄, the absorption edge was located at approximately 460 nm, which equates to a bandgap energy of 2.70 eV. The fact that the CNGOCD aerogel exhibits an absorption edge analogous to that of the pure g‐C₃N₄, indicating that its intrinsic bandgap arises from electron transitions within the VB to CB framework of g‐C₃N₄. Notably, a pronounced increase in background absorption is evident across the 300–800 nm visible‐light spectrum. Enhanced absorption is attributed to structural modifications that facilitate multiple light reflections, a recognized strategy to boost light‐harvesting efficiency in photocatalysts [51]. Improved light absorption capacity can promote the generation of electron‐hole pairs, thereby enhancing catalytic performance. Bandgap energies (E_g) of different samples were calculated using the intercepts of curve tangents with the x‐axis (Figure 8g). Compared to the 2.70 eV bandgap of pure g‐C₃N₄ [38], the 4CN1GO2CD, 4CN2GO2CD and 4CN3GO2CD aerogels exhibited the lowest E_g of 2.02, 2.30, and 2.15 eV after incorporating GO and β‐CD, reflecting strengthened visible‐light responsiveness. For n‐type semiconductors, the conduction band potential (ECB) is set at −0.2 eV relative to EFB. Corresponding valence band potentials (EVB) were determined using Equation (9):

$$E_{VB}=E_{CB}+E_g$$ (9)

FIGURE 8.

Electronic band structure and optical properties of CNGOCD aerogels. (a–f) Mott‐Schottky plots and (g) Tauc plots derived from UV–vis DRS spectra, with the bandgap energies (E_g) estimated from the intercept of the tangent line with the x‐axis. The corresponding band values and structures of CNGOCD aerogels were shown in (h).

Based on bandgap values from Mott‐Schottky plots, the E_VB values for 4CN1GO series aerogels, 4CN2GO2CD, and 4CN3GO2CD were 2.09, 2.33, 2.57, 2.02, 2.30, and 2.15 eV (vs. NHE), respectively (Figure 8h).

3.4.2. Degradation Pathway

To investigate the reactive oxygen species (ROS) involved in photocatalysis, radical scavengers, including isopropyl alcohol (IPA, •OH trap), benzoquinone (BQ, •O₂ ⁻ trap), and ethylenediaminetetraacetic acid (EDTA, h ⁺ trap), were systematically applied [52] (Figure 9a). Intriguingly, RhB removal remained nearly unchanged across all scavenger treatments, likely due to the aerogel's robust pre‐adsorption capacity for RhB, which masked photocatalytic contributions during kinetic measurements. For MB and NR, however, scavenger experiments revealed distinct ROS roles. MB degradation efficiencies declined significantly in both IPA‐ and EDTA‐added solutions, indicating concurrent participation of •OH and h ⁺ in its degradation. The most pronounced reduction occurred with EDTA (50% MB, 47% NR), directly evidencing the dominant role of photogenerated holes in dye oxidation. Conversely, IPA addition caused moderate decreases (74.6% MB, 61% NR), highlighting •OH as a secondary oxidant. Notably, BQ treatment had minimal impact on NR removal (94% efficiency retained), suggesting limited •O₂ ⁻ involvement in this system. These trends establish a ROS hierarchy for dye degradation: h ⁺ > •OH > •O₂ ⁻ . The differential responses between RhB and the other dyes underscore the importance of distinguishing adsorption‐driven removal from photocatalytic processes, particularly when materials exhibit dual‐functionality.

FIGURE 9.

Mechanistic investigation of the photocatalytic degradation process. (a) Effect of various quenchers (IPA for •OH, BQ for •O₂ ⁻, EDTA for h⁺ ) on the photocatalytic degradation efficiency of RhB, MB, and NR, quencher concentration (0.5 mg/mL). (b–e) XPS spectra of 4CN3GO2CD aerogel before adsorption, after equilibrium adsorption of RhB, and after visible‐light photodegradation. Error bars represent SD from three replicate measurements (n = 3).

Mechanistic insights into the adsorption‐degradation of cationic dyes by CNGOCD aerogels were elucidated through pre‐ and post‐reaction XPS spectroscopy (Figure 9b–e). Upon RhB adsorption, enhancements in the N─C═N, O─C═O, and C─O moieties were observed in the XPS spectrum, indicative of dye molecule attachment to the material surface. After the adsorption of RhB, the binding energy of C─O slight increased (from 533.1 to 532.7 eV, Figure 9d), which may be attributed to the formation of hydrogen bonding between the C─O on the aerogel and the C─H on RhB. Following visible‐light irradiation, XPS spectra revealed a spectral profile analogous to the pre‐adsorption state, demonstrating successful degradation and removal of adsorbed RhB and subsequent regeneration of the aerogel to its original condition. Collectively, these results validated the dual‐function capability of the CNGOCD aerogel, integrating both adsorption and degradation processes for efficient dye elimination.

To further confirm that the dye removal proceeds beyond decolorization to genuine mineralization, the TOC content of the RhB solution was monitored over 6 h of visible‐light irradiation. The TOC removal efficiency increased steadily from 56.3% at 1 h to 82.6% at 6 h (Figure S10), demonstrating progressive mineralization of the dye. The TOC removal at early stages lagged behind the dye decolorization efficiency, which is consistent with the stepwise degradation pathway: chromophore cleavage occurs rapidly, while subsequent ring‐opening and oxidation of small organic acid intermediates to CO₂ constitutes the rate‐limiting mineralization step. The eventual attainment of 82.6% TOC removal under prolonged irradiation confirms the deep oxidation capability of the CNGOCD aerogel.

3.4.3. DFT Calculations

Surface electrostatic potential (ESP) provides an effective means for mapping charge distributions and assessing the long‐range electrophilic/nucleophilic reactivity of reacting molecules [53]. Areas displaying higher positive (negative) electrostatic potential are hypothesized to attract nucleophilic (electrophilic) reagent attack preferentially. ESP mappings for GO and g‐C₃N₄ (Figure S11a,d) revealed that the g‐C₃N₄ surface features broad negatively charged regions. These interact strongly with the positively charged domains on GO through electrostatic affinity, H‐bonding networks, and π–π conjugation effects. Such intermolecular forces drive the self‐assembly process into aerogels under hydrothermal synthesis conditions. Fukui index calculations serve to predict the chemical reactions during aerogel synthesis, providing insights into the electron transfer capacities of individual atomic sites within these molecules [54]. Based on reaction theoretical mechanisms and previous reports [35], the ─OH of β‐CD are hypothesized to undergo reactions with the carboxyl moieties in GO. In nucleophilic processes, carboxyl and hydroxyl functional groups are recognized as reactive centers, where Fukui index values are instrumental in identifying atomic sites with the highest susceptibility to nucleophilic attack—positions with elevated Fukui indices tend to engage more readily with nucleophiles due to favorable electron density distributions and orbital polarization. DFT calculation results demonstrated that the reactive sites of GO were predominantly localized on its carboxyl groups, whereas those of β‐CD were concentrated on its hydroxyl groups (Figure S11b,c,h,i). This evidence further supported the hypothesis that β‐CD hydroxyl groups form chemical bonds with GO molecules through nucleophilic reactions, thereby enhancing the structural stability of the aerogel.

DFT calculations were further employed to mechanistically clarify the differential adsorption‐degradation performances of CNGOCD aerogels toward three cationic dyes (Figure 10). ESP mappings and extremum point distributions revealed that RhB exhibited uniformly distributed positive potential maxima around its molecular structure, with a higher surface positive charge density compared to MB and NR. Electrostatic feature explained the superior adsorption capacity of aerogels for RhB (Figure 10g–i). According to frontier molecular orbital (FMO) theory, the highest occupied molecular orbital (HOMO) correlates with electrophilic reactivity, while the lowest unoccupied molecular orbital (LUMO) relates to nucleophilic processes [55]. HOMO‐LUMO energy gap dictates molecular chemical stability and optical properties. Smaller gaps indicate greater reactivity. DFT calculations results showed RhB possessed the narrowest HOMO‐LUMO gap (ΔE = 2.15 eV), rendering it most amenable to degradation. In contrast, MB exhibited the largest energy gap (ΔE = 3.74 eV), accounting for its highest degradation resistance among the three dyes (Figure 10a–f).

FIGURE 10.

DFT‐calculated frontier molecular orbitals and ESP maps of the three target dyes. HOMO‐LUMO and surface electrostatic potential of RhB (a, d, g), MB (b, e, h), and NR (c, f, i). (g–i): the green: the negative electrostatic potential; the blue: the positive part; the cyan ball: the minimum electrostatic potential; the yellow ball: the maximum electrostatic potential.

4. Conclusion

In this study, a metal‐free ternary composite aerogel (g‐C₃N₄/GO/β‐CD) was developed by an eco‐friendly approach, enabling synergistic adsorption and photocatalysis for efficient dye wastewater purification. The engineered aerogel exhibited exceptional performance, achieving a maximum adsorption capacity of 653.2 mg/g and a 97.1% removal efficiency for RhB. This outperformed binary composites and retained functionality across five recycling cycles without notable performance decline. Mechanistic analysis revealed key structural and electronic enhancements: (1) GO facilitated charge separation in g‐C₃N₄ by reducing electron‐hole recombination, thereby boosting photocatalytic activity; and (2) β‐CD disrupted nanosheet stacking between GO and g‐C₃N₄ layers, enhancing structural integrity and exposing additional active sites for dye adsorption. These design features established a robust correlation between material architecture and functional performance, enabling efficient integration of adsorption processes with photocatalytic degradation. The developed aerogel presents a promising solution for real‐world wastewater treatment challenges, offering novel design guidelines for next‐generation composite materials that balance efficiency, stability, and ecological compatibility. Beyond organic dyes, the established multi‐modal binding mechanisms, such as electrostatic attraction, hydrogen bond interaction, and hydrophobic host‐guest inclusion, render the CNGOCD aerogel a promising candidate for removing other cationic and hydrophobic micropollutants, such as fluoroquinolone antibiotics and phenolic endocrine disruptors, which merit dedicated investigation in future studies.

Author Contributions

Yuan Yu: conceptualization, methodology, investigation, data curation, visualization, and writing – original draft. Ke Zhou: methodology, investigation, writing – review and editing. Lu Shang: methodology, investigation, writing – review, and editing. Yuting Zou: methodology and investigation. Qing Yin: methodology and investigation. Shengzhe Wang: methodology, writing – review, and editing. Na Chang: methodology, writing – review, and editing. Yuting Wen: conceptualization, methodology, resources, funding acquisition, supervision, project administration, writing – review, and editing. Jun Li: conceptualization, methodology, resources, funding acquisition, supervision, project administration, writing – review, and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting file: smll74507‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that supports the findings of this study are available in the supplementary material of this article.