Biomimetic aerogels with hierarchical honeycomb architecture for superior CO₂ adsorption, selectivity, and structural integrity

Abstract

In structured adsorbents, achieving mesoporosity, crucial for efficient gas adosorption, is challenging, which restricts mass transport and accessibility to active sites. Here, we address this limitation by developing the first hierarchically porous honeycomb aerogels that replicate hexagonal pores at both the macro-level and micro-level wall structure. This design, inspired by nature’s most efficient patterns, enables us to achieve CO₂ adsorption capacity (3.94 mmol g⁻¹ at 298 K and 1 bar), selectivity (65.2 CO₂/N₂), and high specific surface area (370 m² g⁻¹). The honeycomb aerogels are constructed from manganese dioxide (MnO₂) functionalized electrochemically exfoliated graphene (MEEG) and chitosan (CS). By optimizing the MnO₂ loading and the MEEG to CS weight ratio, we achieved dual-scale hexagonal porosity, enabling a hybrid physical and chemical adsorption mechanism. The hybrid adsorption leverages the rapid kinetics of chemisorption and ease of regeneration characteristic of physisorption, making these materials highly efficient. This highlights the synergy between enhanced surface accessibility of primary amine groups and selective adsorption properties, setting a new standard for hierarchically structured materials.

Honeycomb structures maximize surface area, important for gas adsorption, but replicating them synthetically is challenging. Here, an aerogel with hexagonal porosity was synthesized with MnO₂-functionalized electrochemically exfoliated graphene for carbon dioxide adsorption and high selectivity over nitrogen

Introduction

Hierarchical honeycomb structures are among nature’s most efficient and multifunctional designs^1,2. In diatoms, these silica-based honeycomb structures maximize surface area, supporting nutrient uptake and CO₂ absorption essential for photosynthesis³. Coral polyps exhibit multi-scale honeycomb formations that facilitate water flow and nutrient transport to symbiotic algae, significantly enhancing photosynthetic efficiency⁴. The lightweight yet strong honeycomb skeleton in coral also provides resilience against ocean currents. Bones similarly rely on these structures, with a dense outer layer (cortical bone) and inner spongy matrix (trabecular bone) combining strength, lightweight properties, and adaptability to support body weight and absorb impacts⁵. These natural examples show how hierarchical honeycomb structures enable optimal functionalities, inspiring engineered materials for mass transport-limited applications, such as carbon capture, energy storage, catalysis, and robust, lightweight construction materials⁶.

However, achieving the replica of these hierarchically porous honeycomb structures through synthetic methods has proven challenging. While techniques like templating, freeze-casting, and 3D printing were applied successfully to create single-scale honeycomb patterns^2,7–11, they have failed to achieve hexagonal architecture across both macro- and micro-levels. 3D printing is effective for producing regular honeycomb patterns but typically lacks the resolution to fully replicate the fine, layered details and nanoscale intricacies observed in structures like diatoms or coral^6,7. Additionally, controlling material composition to achieve multifunctionality (e.g., both mass transfer enhancement, accessibility of the active sites, and lightweight strength) of natural counterparts remains challenging. As a result, synthetic approaches yield simpler or scaled-down versions that, while functional, do not fully reach the performance potential of nature’s hierarchical honeycomb structure^5–7.

Herein, we report the first aerogel obtained via colloidal assembly that achieves dual-scale hexagonal porosity, defined as the presence of hexagonal pores at both the macro-level (visible at micrometer scale) and the micro-level (with nanoscale intricacies within the pore walls), mimicking the hierarchical honeycomb morphology found in nature. These aerogels are derived by synthesizing MnO₂-functionalized electrochemically exfoliated graphene (MEEG) and crosslinking it with chitosan (CS) through amide covalent bonding using EDC-NHS (1−ethyl−3−(3−dimethylaminopropyl)carbodiimide hydrochloride and N−hydroxysuccinimide) as a crosslinker. By adjusting composition of the precursor colloidal system, we demonstrate the tunability of the freeze-dried aerogel’s structure, ranging from single-scale honeycomb patterns to fully hierarchical honeycombs with hexagonal pores at both the macro-level and micro-level wall structure. As a model mass-transport limited application, we test our architecture for CO₂ adsorption. Our findings highlight the effectiveness of hierarchical honeycomb structures, revealing record-breaking CO₂ adsorption capacity in structured adsorbents. In addition, the aerogel displays rapid adsorption kinetics through a hybrid mechanism of both physical and chemical adsorption, enabled by dual-scale porosity and an approximately threefold increase in surface area compared to macro-scale honeycomb structure. Under optimal composition, the hierarchically structured aerogels displayed a surface area of 374.2 m² g⁻¹, a corresponding CO₂ adsorption capacity of 3.94 mmol g⁻¹, and a selectivity of 65.2 at 298 K and 1 bar. Notably, this aerogel maintains 98.9% of its adsorption capacity after 20 regeneration cycles demonstrating exceptional structural integrity. Additionally, these aerogels exhibited an electrical conductivity of 18.4 ± 0.2 S m⁻¹, enabling direct electrical heating of the sorbent for regeneration and reduced energy costs. This design opens new possibilities for mass transport-limited applications, offering promising advancements in areas such as CO₂ capture, energy storage, and catalysis.

Results and discussion

Mechanistic design of MEEG-CS aerogels

The hierarchical honeycomb aerogel developed in this study exhibits microstructural similarities to architectures occurring in nature (bees’ honeycomb, coral polyps, silica-based diatoms, and human bone) are shown in Fig. 1a. A thorough analysis to find the optimal MEEG: CS ratio in the 1 wt.% and 2 wt.% concentration of the precursor gel was performed (see Supplementary Information Section S1, Fig. S1–S4). The analysis in the supplementary information reveals that the optimal MEEG: CS ratio is 1:1 at 1 wt.% total solids, providing the highest BET surface area and CO₂ adsorption capacity (3.94 mmol g⁻¹). Increasing the MEEG content enhances crosslinking, surface area, and pore accessibility, but excessively high ratios (e.g., 3:2) compromise structural integrity. Lower loadings (<1 wt.%) were also investigated but failed to produce stable aerogels due to insufficient structuring and poor gelation, resulting in collapse during drying. Higher loadings (>2 wt.%) led to aggregation and phase separation, adversely affecting material uniformity and mechanical stability. The 1 wt.% aerogel outperforms the 2 wt.% variant due to its less dense structure, facilitating better surface area utilization despite fewer active adsorption sites. The CO₂ adsorption in MEEG-CS aerogels arises from the synergy between physisorption, driven by high surface area and porous structure, and chemisorption, facilitated by strong interactions between CO₂ and -NH₂ groups. This combination ensures high capacity, rapid kinetics, and cost-efficient performance. The optimized 1 wt.% MEEG-CS aerogel with a 1:1 MEEG: CS ratio was used for all further analyses.

Fig. 1. Biomimetic design of hierarchical honeycomb aerogels.

a Comparison of MEEG-32-CS microstructure with several hierarchical honeycomb structures observed in nature - beeswax, coral polyps, silica-based diatoms, and human bone microstructure. b SEM images of MEEG-CS aerogels, as a function of MnO₂ loading% in MEEG. c Thermogravimetric analysis of the four MEEG samples under air. d FTIR spectra for EEG, MEEG-12, MEEG-24, MEEG-32, and MEEG-42. The main characteristic bands are shown in the figure. e BET surface area and Porosity of MEEG-CS aerogels as a function of MnO₂ loading% in MEEG.

By varying the amount of MnSO₄ concentration during the electrochemical exfoliation process (7.5, 10, 20 and 30 mM), the MnO₂ loading and consequently the hierarchical architecture was optimized, as depicted by the four samples shown in Fig. 1b. To determine the mass loading of MnO₂ in the four MEEG samples prepared°C TGA was performed under air (°C) reported in Fig. 1c. By increasing the temperature from 25 °C to 900 °C in the air environment, the graphene was decomposed, while MnO₂ was converted into Mn₂O₃¹². The residual mass percentages of 10.9 wt.%, 22.1 wt.%, 29.2 wt.% and 37.9 wt.% correspond to the remaining Mn₂O₃, which is equivalent to MnO₂ mass loadings of 12 wt.%, 24.3 wt.%, 32.1 wt.% and 41.7 wt.%, respectively. These MEEG samples are denoted as MEEG-12, MEEG-24, MEEG-32, and MEEG-42 from here onwards.

Figure 1d depicts the FTIR spectra of the control EEG (electrochemically exfoliated graphene, without MnO₂ functionalization) and MEEG samples. The spectrum of EEG exhibited an absorption peak at 1565–1580 cm⁻¹, corresponding to the C=C bonds present in the graphene sheets. Additionally, a peak at 1220-1230 cm⁻¹ was observed, which was attributed to the presence of carboxylic groups¹³. Following the intercalation of MnO₂ between the MEEG sheets, a new prominent peak emerged at 550–560 cm⁻¹, indicating the presence of Mn-O bonds in MnO₂¹³. This confirmed the successful functionalization of MEEG with MnO₂ nanoparticles. By increasing MnO₂ content from 12 wt.% in MEEG-12 to 41.7% in MEEG-42, 1220–1230 cm⁻¹ peak corresponding to the carboxylic groups becomes weaker and almost disappears for MEEG-42. However, simultaneously the MnO₂ band becomes increasingly prominent, thus proving that MnO₂ was introduced at the expense of -COOH groups, which are primarily responsible for crosslinking with CS (see Supplementary Information Section S2, Fig. S5).

The surface area enhancement (Fig. 1e), as expected, is intricately linked to the obtained morphology in the precursor hydrogel as well as the microstructure of the derived aerogels of MEEG-CS systems, with a maximum of 374.2 m² g⁻¹ at an MnO₂ loading of 32 wt.% (Figs. S6, and 1b). Figure S7a demonstrates that N₂ adsorption-desorption isotherms of the MEEG-32-CS sample at 77 K exhibited a small reversible hysteresis loop. Thus, the sample depicted a typical IUPAC type IV isotherm characteristic, indicating an open hierarchical structure with more mesopores (2–50 nm) than micropores (<2 nm). The pore size distribution profile (Fig. S7b) also shows that the MEEG-32-CS aerogel has pores with diameters ranging from 20 to 500 Å. The total pore volume at P/P₀ = 0.99 was determined to be 0.34 cc g⁻¹ (Table S1). Henceforth, MEEG refers to MEEG-32, unless otherwise specified.

Increasing the surface area in graphene-based systems is often hindered by the formation of aggregates at higher particle loadings caused by π-π stacking between consecutive graphene sheets. In contrast, the MEEG-CS crosslinked system (Fig. 2a) demonstrated a significant two-fold increase in surface area compared to the EEG-CS system without MnO₂ functionalization or crosslinking (Fig. 2a, b and Table S1). The presence of MnO₂ impedes the proximity and subsequent π-π stacking of graphene sheets¹³. This is achieved by the MnO₂ particles preventing close stacking through steric hindrance, as well as increasing the electrostatic repulsion between the graphene sheets. The enhanced zeta potential values (EEG: –33.2 ± 0.8 mV; MEEG: –36.7 ± 1.3 mV) confirmed the increase in electrostatic repulsion due to MnO₂ functionalization, resulting in better dispersion and reduced aggregation. In addition, the negatively charged oxygen atoms in MnO₂ particles on the graphene surface enhance electrostatic stabilization^14–16. As a result of this suppressed MEEG aggregation, the increase in surface area from 52.8 m² g⁻¹ for EEG-CS aerogel (without crosslinking) to 103.6 m² g⁻¹ in MEEG-CS aerogel (without crosslinking) was observed. On the other hand, the crosslinking process forming strong amide bonds between MEEG and CS played a crucial role in inhibiting the growth of ice crystals in the preferred direction during aerogel synthesis, resulting in hierarchically structured honeycombs, after freeze-drying¹⁷. This structure is further evidenced by SEM micrographs (Figs. 2b, and S3a), which reveal well-organized honeycomb frameworks at both 20 μm and 1 μm scales. These observations confirm the presence of dual-scale, hierarchically porous architectures that contribute to enhanced diffusion and surface accessibility. Crosslinking between MEEG and CS using EDC-NHS significantly contributed to the formation of mesoporous composites by enhancing intermolecular interactions and structural integrity during synthesis. The zeta potential further improved after crosslinking (–40.3 ± 1.2 mV), consistent with the observed increase in surface area. The crosslinking reaction, facilitated by 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride and N-hydroxysuccinimide (EDC-NHS), involved the activation of carboxyl groups (-COOH) on the MEEG surface by EDC. This reaction leads to the formation of an unstable O-acylisourea intermediate, which reacts with the NHS (N-hydroxysuccinimide) to form a semi-stable NHS ester intermediate (Fig. 2a)^18–20. The intermediate NHS ester undergoes a reaction with the amine groups present on CS, leading to the formation of covalent amide bonds, as illustrated in Fig. 2c. In addition to covalent bonding, MEEG can also establish hydrogen bonds with CS and engage in weak CH-π and heteroatom-π interactions^21–23. The EDC-NHS crosslinking method offers advantages over other crosslinkers, such as glutaraldehyde and epichlorohydrin, as it enables controlled and reproducible crosslinking of carboxyl groups in MEEG and amino groups in CS, reducing side reactions and the need for harsh conditions or toxic chemicals, making it environmentally friendly and easily scalable (see Supplementary Information Section S2, Figure S5)^24,25. Upon crosslinking, the surface area increases to 220.4 m² g⁻¹ for the EEG-CS system and to 374.23 m² g⁻¹ for the MEEG-CS system (Fig. 2d).

Fig. 2. MEEG-CS crosslinking and MnO₂ functionalization enable a two-fold increase in surface area.

a Overall reaction and crosslinking mechanism between MnO₂ functionalized electrochemically exfoliated graphene (MEEG) and chitosan (CS). b SEM micrographs of EEG-CS aerogel without crosslinking; MEEG-CS aerogel without crosslinking; EEG-CS aerogel with crosslinking and MEEG-CS aerogel with crosslinking. c FTIR spectra for MEEG, CS, and MEEG-CS aerogel. The main characteristic bands are shown in the figure. d BET surface area and Porosity (%) of the derived aerogels. e Direct correlation observed between the BET surface area and CO₂ uptake for the derived aerogels. The yellow category refers to the crosslinked EEG-CS aerogels, and the green category refers to the crosslinked MEEG-CS aerogels.

The correlation between CO₂ adsorption capacity and the surface area (Fig. 2e) underscores the critical role of the aerogel architecture. EEG-CS aerogels exhibit a traditional honeycomb microstructure, whereas MEEG-CS aerogels transition to a hierarchical honeycomb structure, accompanied by a significant increase in surface area. In this architecture, the macro-level hexagonal pores (~10–15 µm) serve as primary transport channels that enable rapid CO₂ diffusion into the aerogel. The micro-level wall structures, composed of interconnected nanosheets and pores (<1 µm), offer an extensive surface for adsorption and facilitate capillary condensation. This enhancement in the surface area provides more sites for CO₂ interaction, enhancing physisorption, while primary amine sites on the surface facilitate chemisorption. The synergy between these mechanisms significantly enhances CO₂ uptake. These insights are crucial for the development of efficient and sustainable carbon capture materials. The CO₂ adsorption capacity of the MEEG-CS adsorbent is constrained by a balance between the availability of carboxyl groups for crosslinking and the MnO₂ content required to prevent MEEG aggregation. While introducing additional carboxylic group defects can potentially enhance surface area, it necessitates incorporating MnO₂ as a stabilizing barrier. This compromise limits further crosslinking enhancement, highlighting the delicate synergy between crosslinking and stabilization achieved in this study.

Adsorption isotherms and thermal regeneration via electrical heating

The hierarchical structure of the aerogels improves adsorption kinetics by reducing mass-transfer resistance and increasing accessibility to active adsorption sites. Among the derived aerogels, MEEG-CS exhibits a maximum CO₂ adsorption capacity of 3.94 mmol g⁻¹ (Fig. 3a). Consequently, the increase in CO₂ adsorption capacity from 1.13 mmol g⁻¹ for EEG-CS aerogel (without crosslinking) to 1.74 mmol g⁻¹ in MEEG-CS aerogel (without crosslinking) was observed (Fig. 3b). With crosslinking, the adsorption capacity is also correspondingly increased to 3.52 mmol g⁻¹ for the EEG-CS system and to 3.94 mmol g⁻¹ for the MEEG-CS system, respectively (Fig. 3b). Despite losing some primary amine sites during the crosslinking process, the primary objective of increasing the surface area to enhance physisorption was effectively accomplished.

Fig. 3. Regenerability, stability, selectivity, and CO₂ capture performance of MEEG-CS aerogels.

a CO₂ breakthrough curve for MEEG-CS crosslinked aerogels with varying loading of MnO₂ on MEEG, at 298 K and 1 bar. b CO₂ breakthrough curves for the derived aerogels- EEG-CS without crosslinking; MEEG-CS without crosslinking; EEG-CS with crosslinking and MEEG-CS with crosslinking at 298 K and 1 bar. c CO₂ and N₂ isotherms at 273 K, 288 K, 298 K, and 308 K on MEEG-CS crosslinked aerogel. d CO₂/N₂ selectivity of MEEG-CS crosslinked aerogel calculated from pure adsorption isotherms using the IAST model, as a function of pressure. e Regeneration efficiency over adsorption and regeneration via. electrical heating cycles for 20 cycles for MEEG-CS crosslinked aerogel. f SEM images of the MEEG-CS crosslinked aerogel took before and after 20 adsorption-regeneration via. electrical heating cycles. The top row for each case has a scale bar of 20 μm while the bottom row has a scale bar of 1 μm.

Pure CO₂ and N₂ isotherm analysis was conducted on the MEEG-CS crosslinked aerogel at four different temperatures (273, 288, 298, and 308 K), within a pressure range of 0–1200 mbar (Fig. 3c). The MEEG-CS crosslinked aerogel exhibits a synergistic combination of physisorption and chemisorption mechanisms. Chemisorption, driven by basic amine groups forming carbamates, dominates at low pressures (<200 mbar), while physisorption, reliant on mesoporous structure and van der Waals interactions, becomes significant at higher pressures (>200 mbar). This dual behavior enhances CO₂ uptake across a wide pressure range. The decreasing nature of CO₂ uptake capacity from 5.43 mmol g⁻¹ at 273 K to 3.87 mmol g⁻¹ at 308 K suggested that the adsorption is highly thermodynamically controlled, which is a defining characteristic of physisorption²⁶. However, Fig. 3c also illustrates that despite having a mesoporous structure, the CO₂ uptake of MEEG-CS crosslinked aerogel is significant (>2 mmol g⁻¹) even in the lower-pressure regions (below 200 mbar), indicating that the adsorption is governed by chemisorption. The high CO₂ adsorption capacities observed in the low-pressure range can be attributed to the abundance of basic amine groups, which serve as active sites for chemical adsorption, resulting in carbamate formation²⁷. The presence of amine functional groups on the surface of the aerogel contributes significantly to the enhancement of CO₂ uptake capacity, especially at lower pressures where CO₂ interaction with basic NH₂ groups is more significant than physisorption associated with the pore volume and specific surface area. With increasing pressure, the mesopores aid in accessibility to the active adsorption sites, thus enhancing the overall CO₂ uptake. The synergistic interaction between physisorption and chemisorption plays a critical role in boosting both the adsorption efficiency and selectivity of the aerogel. At low pressures, chemisorption sites selectively bind CO₂ molecules through the formation of strong carbamate linkages, ensuring that the initial capture is rapid and highly specific. As the pressure increases, the porous network provides additional physical adsorption sites, allowing more CO₂ to be accommodated via weaker van der Waals forces. This complementary behavior ensures a continuous uptake profile rather than a saturation plateau, effectively extending the material’s working pressure range. Moreover, the physical pores assist in diffusive transport, enabling more CO₂ molecules to access internal chemisorption sites. This cooperative mechanism not only maximizes the utilization of available binding sites but also mitigates desorption at intermediate pressures, resulting in enhanced retention and lower regeneration energy requirements. The integration of both mechanisms thereby leads to superior performance in practical CO₂ capture scenarios, especially under variable operating conditions such as those in industrial post-combustion processes. The isotherms thus reflect a truly synergistic combination of physisorption and chemisorption mechanisms that collectively contribute to the high CO₂ uptake capacity and efficient pressure-responsive behavior of the MEEG-CS crosslinked aerogels.

In contrast, the N₂ uptake is low due to the lack of electrostatic interactions between the N₂ molecules, and both the basic amine functional groups and the introduced MnO₂ functional groups on the MEEG-CS crosslinked aerogel. However, due to the presence of mesopores the N₂ adsorption is enhanced marginally with an increase in pressure. The CO₂ heat of adsorption value calculated using the Clausius-Clapeyron equation from the four adsorption isotherms of MEEG-CS crosslinked aerogel was −49.2 kJ mol⁻¹. The CO₂ heat of adsorption value of −49.2 kJ mol⁻¹ for the MEEG-CS crosslinked aerogel aligns with the synergistic combination of physisorption and chemisorption mechanisms. Physisorption, which is dominated by van der Waals interactions, typically exhibits heat of adsorption values ranging from −10 to −40 kJ mol⁻¹. In contrast, chemisorption, involving the formation of chemical bonds (such as carbamate formation with amine groups), is characterized by higher heat of adsorption values, typically ranging from −80 to −250 kJ mol⁻¹. The observed heat of adsorption of -49.2 kJ mol⁻¹ reflects the contribution of both mechanisms, with chemisorption playing a dominant role at lower pressures and physisorption enhancing adsorption at higher pressures. The experimental data obtained from CO₂ and N₂ adsorption were fitted to Freundlich and Toth isotherm models (Fig. S8 and Table S2). The correlation coefficients of the Toth isotherm model suggest that the adsorption takes place on a heterogeneous surface²⁸. The parameters obtained from the Freundlich isotherm model imply that the adsorption strength of CO₂ on the surface of MEEG-CS crosslinked aerogel was non-uniform, with a n value that deviates from 1, indicating multi-layer adsorption on a non-uniform surface²⁹. The selectivity shows a decreasing trend with increasing pressure (Fig. 3d). The maximum selectivity is 174.9 at 4 mbar, decreasing to 65.2 at 1 bar, which is consistent with the characteristics of porous materials.

The resistive heating method used in this study enabled the direct heating of the aerogel, made possible due to the electrical conductivity of MEEG^30–35. This overcomes the need for external heating procedures, resulting in a decrease in energy usage as the MEEG-CS crosslinked aerogel was heated in seconds instead of several minutes in external heating systems. The surface morphology of the aerogel remained intact after 20 adsorption and electrical heating regeneration cycles, retaining 98.9% of its initial adsorption capacity (Fig. 3e). The SEM images of adsorbents confirmed minimal changes in the surface morphology due to electrical heating regeneration (Fig. 3f). Cyclic pressure swing adsorption (PSA) studies also demonstrate the long-term stability of the sorbent (Fig. S9).

The water sorption profile for MEEG-CS crosslinked aerogel is depicted in Fig. S10. At a relative humidity of 100%, the water uptake was only 7.8 ± 0.4 mmol g⁻¹, which confirms that the final composite underwent partial hydrophobization during crosslinking with hydrophobic CS at 60 °C and freeze-drying. Interestingly, the water uptake of MEEG-CS was significantly lower than Zeolite 13X, as well as three other water-resistant adsorbents, namely CAU-10, CALF-20, and Al fumarate^36–39. Additionally, exposure to 100% relative humidity did not disrupt the microstructure of the aerogel (Fig. S11), making it a promising candidate for humid flue gas applications.

In Fig. S12, a comparison of CO₂ uptake capacities and CO₂/N₂ selectivity of comparable hierarchically structured adsorbents reported in recent research articles is presented^40–56. The MEEG-CS crosslinked aerogel shows an impressive, unprecedented balance between selectivity and adsorption capacity at 298 K and 1 bar, compared to other hierarchically structured CO₂ adsorbents. Although several non-structured CO₂ adsorbents (e.g., Zeolite-13X, CALF-20, COFs, etc.) have shown better adsorption capacity (e.g., Zeolite-13X with a CO₂ uptake of 5.1 mmol g⁻¹ at 298 K and 1 bar), hierarchically structured adsorbents such as MEEG-CS aerogels have the advantage of reduced pressure drop across the adsorption columns during operation. This results in reduced operational costs, making them an attractive solution for large-scale systems^35,38. Hybrid silica aerogels have shown a CO₂ adsorption capacity of 4.9 mmol g⁻¹ at 273 K and 1 bar, coupled with a surface area of 756 m² g⁻¹⁵⁷. Remarkably, MEEG-CS exhibits a superior adsorption capacity (5.43 mmol g⁻¹ at 273 K and 1 bar), despite possessing a surface area of merely half that of the silica aerogel, 374.2 m² g⁻¹. We attribute this enhanced performance to its unique hierarchical honeycomb microstructure^58–61. This comparison highlights the trade-offs between surface area and adsorption capacity, as higher surface area adsorbents often exhibit higher CO₂ uptake but may also suffer from higher pressure drops, which hinder the efficiency of gas flow through the adsorbent material. These adsorbents may also have lower selectivity, making them less effective in environments with a mixture of gases. In contrast, MEEG-CS aerogels, with a comparatively lower surface area, are able to achieve a favorable balance between adsorption capacity, CO₂/N₂ selectivity, and lower pressure drop, thus providing a more effective and cost-efficient solution for large-scale CO₂ capture. Furthermore, the hybrid adsorption mechanism in MEEG-CS enhances its performance, enabling it to capture CO₂ efficiently even with a lower surface area compared to traditional materials. Another advantage of the derived aerogels is that they can be regenerated through a simple Joule heating process facilitated by the conductivity imparted by MEEG. Joule heating regeneration for CO₂ sorbents is a comparatively innovative approach, and only a limited number of studies have explored this method to date^30–35,56. A comparison study of adsorbents regenerated using the Joule heating methodology is presented in Table S3. The MEEG-CS aerogels demonstrate competitive performance in CO₂ uptake (3.96 mmol g⁻¹), CO₂/N₂ selectivity (65.2), and high regeneration efficiency (99%) over 20 cycles, comparable to other advanced adsorbents like CNT films and CNT-zeolite^30–32. Despite a lower BET surface area (374.2 m² g⁻¹), the aerogels exhibit strong adsorption and regeneration capabilities due to their structured hierarchical honeycomb architecture, making them a promising candidate for CO₂ capture using electrical heating. MEEG-CS aerogels show promising results regarding CO₂ adsorption capacity due to hybrid adsorption mechanism, CO₂/N₂ selectivity, operational costs, adsorption kinetics, and regeneration stability, with additional characteristics of simple synthesis and a sustainable approach.

Conclusions

In this study, MEEG-CS aerogels were successfully synthesized using a crosslinking approach to develop hierarchically porous honeycomb structures with dual-scale hexagonal porosity. This biomimetic design, inspired by nature’s efficient patterns, allowed for the achievement of remarkable CO₂ capture performance, including an adsorption capacity of 3.94 mmol g⁻¹ at 298 K and 1 bar, exceptional CO₂/N₂ selectivity (65.2), and a specific surface area of 370 m² g⁻¹. The functionalization of EEG with MnO₂was employed to prevent aggregation, resulting in enhanced surface accessibility and adsorption efficiency. The synergy between physisorption and chemisorption mechanisms was facilitated by optimized surface chemistry and hierarchical porosity, highlighting the potential of MEEG-CS aerogels as a sustainable and effective solution for CO₂ capture and regeneration. Exceptional structural integrity was exhibited by the aerogels, along with long-term stability over 20 regeneration cycles and resilience in humid environments without requiring additional reduction steps. A two-fold enhancement in aerogel surface area through surface chemistry optimization was reported for the first time, providing new insights into the design of advanced CO₂ adsorbents. A benchmark has been set for the development of hierarchically structured materials, offering a pathway toward scalable and commercially viable solutions for CO₂ capture.

Experimental section

Materials

Graphite foil (21 cm², 0.5 mm thick, 99% pure), manganese sulfate (Mn)₂SO₄, ammonium sulfate (NH₄)₂SO₄, EDC (1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride and NHS (N-hydroxysuccinimide) and chitosan (98% deacetylated, from shrimp, M_w <100 kDa, used without further purification), were purchased from Millipore Sigma. Glacial acetic acid (CH₃COOH) was purchased from Fisher Scientific.

Synthesis of MEEG

Electrochemical exfoliation of graphite inside a mixture of (NH₄)₂SO₄ and MnSO₄ as the electrolyte was used to synthesize the graphene. Graphite and stainless-steel foil were cut in the shape of (4 × 1) rectangular strips as the anode and cathode and immersed inside 40 ml of electrolyte while the distance between them was fixed at 2 cm. The concentration of (NH₄)₂SO₄ was fixed at 0.1 M, and four different concentrations of 7.5, 10, 20, and 30 mM were selected for MnSO₄ to prepare the graphene composites with different mass loadings of MnO₂. The electrochemical exfoliation of graphite was performed by applying a constant voltage of 10 V followed by washing the products by a vacuum filtration setup. In the next step, the products were dispersed in deionized water after 1 h of bath sonication and left overnight for the settlement of unexfoliated graphite particles. The supernatants were separated and freeze-dried to prepare the graphene powder (S3, Fig. S13).

Preparation of crosslinked MEEG-CS aerogels

MEEG dispersions (50 mL of 1% and 2%) were prepared in deionized water (>18 M cm, Milli-Q) and probe-sonicated at a frequency of 50 Hz for 30 min. Different ratios of MEEG to chitosan (MEEG-CS) were used to generate hybrid aerogels with solid contents of 1 and 2 wt. %. For instance, MEEG-CS 1 wt.%, 1:1 sample was produced by dispersing 50 mg of MEEG in 5 ml of deionized water and then adding 50 mg of chitosan dispersion in 5 ml of 1% acetic acid solution to it. Five milligrams of EDC-NHS in a 1:3 ratio was gradually added to the resultant solution at 60 °C and was magnetically stirred for 24 h before being poured into a glass vial and aged for 2 days (Detailed optimization of crosslinking time is shown in S4, Fig. S14). The gel was then frozen using liquid nitrogen and placed in a freeze dryer (LABCONCO Benchtop Freeze Dryer, operating conditions: 0.005 mbar and −104 °C) for 48 ± 1 h, resulting in the aerogel. Finally, the adsorbents were thermally treated by keeping them in a vacuum oven at 190 °C for 24 ± 1 h^17,42–44.

Characterization

The Leica SP8 TSC inverted microscope with an 8 kHz resonant scanner and a white laser excitation source was utilized to examine the precursor gels. In the case of MEEG, which exhibits inherent fluorescence in the 500–700 nm range, a 480 nm laser was used to generate an emission spectrum ranging from 520 to 700 nm. CS, on the other hand, does not exhibit fluorescence and was imaged using reflection mode with a 450 nm laser. 2D images of the MEEG-CS aerogels were captured with a 63× magnification objective, 1024 × 1024 image format, 180.38 nm × 180.38 nm pixel size, 0.8 AU pinhole, 184.52 mm × 184.52 mm image size, and 400 Hz scanning speed. Avizo software was used to apply an image processing technique to determine the mean particle diameter. To analyze the pore morphology of the MEEG-CS crosslinked aerogels, the cross-sections were examined by a ZEISS Supra 55-VP field emission scanning electron microscope (SEM) used with Pt pre-deposition. The images were captured using a 10 kV acceleration voltage with 1024 × 1024-pixel resolution, and the samples were mounted on conductive carbon glue tabs on an aluminum stub.

The Perkin Elmer Frontier FT-IR spectrometer was used to collect Fourier-transform infrared spectroscopy (FT-IR) spectra, with a resolution of 4 cm⁻¹ and 32 scans. X-ray diffraction (XRD) analysis was carried out on the aerogels using a Rigaku tabletop diffractometer, which utilized Cu K_a radiation at 40 kV and 25 mA to determine their phase structure and composition. To determine the elemental composition of the synthesized MEEG, X-ray photon spectroscopy (XPS) was performed using Thermo-Scientific’s ESCALAB 250Xi. To determine the moisture uptake, the Cincinnati sub-zero humidifier chamber was employed. The Malvern Instruments’ Zetasizer was used to conduct Zeta potential and Dynamic Light Scattering (DLS) tests.

To examine the viscoelastic properties of the MEEG-CS crosslinked gels, oscillatory frequency sweep measurements were carried out using an Anton Paar Modular Compact Rheometer (MCR 702). The rheological measurements were performed with a 50 mm cone plate geometry having a 0.286 mm gap. The oscillatory frequency was ramped from 0.1 to 100 Hz, while the amplitude was logarithmically ramped between 0.01 and 100 % shear strain, and the measurements were repeated thrice to verify the reproducibility of the results.

To examine the nitrogen adsorption isotherms of the aerogels, a BELSORP-mini was employed at a temperature of 196 °C (MicrotracBEL Corp.). The specific surface area (SBET) was determined using the Brunauer-Emmett-Teller (BET) method in the pressure range of 0.05–0.35. The total pore volume (V_t) was calculated from the quantity of N₂ adsorption at P/P₀ = 0.99. The Barrett-Joyner-Halenda (BJH) approach was used to evaluate the pore size distribution by analyzing the N₂ adsorption-desorption isotherm at 77 K.

The apparent bulk density (ρ_b) of the aerogel was obtained from the aerogel mass (m) and its volume (V) as described in (Eq. 1):

$${\rho }_b=\frac{m}{V}$$ 1

The values of skeletal density (ρ_s) were obtained by using helium in a gas-displacement pycnometer. The porosity of the aerogels was calculated according to the following relation, as described in (Eq. 2):

$$\mathit{Porosity},\%=\left(\mathbf{1}-\frac{{\mathit{\rho }}_{\mathit{b}}}{{\mathit{\rho }}_{\mathit{s}}}\right)*100$$ 2

Through thermogravimetric analysis (TGA) on a TA Instruments High Resolution 2950, the thermal stability of the organic aerogels was assessed. The TGA thermograms were produced using air/nitrogen and a ramp rate of 10 °C min⁻¹ at temperatures between 20 and 900 °C.

CO₂ adsorption tests were conducted on the MEEG-CS crosslinked aerogels using TGA, with a sample weighing around 5 mg subjected to 100 ml min⁻¹ of pure Ar gas for degassing. After retaining the sample under the Ar gas flow, the temperature was lowered to 25 °C, and the feed gas was changed to CO₂ for a 30-min adsorption procedure. The experimental error in the sample’s capacity to trap CO₂ was estimated within 5% using three repeated adsorption tests using the MEEG-CS crosslinked aerogels.

Adsorption isotherms

The adsorption isotherms for CO₂ and N₂ were measured using the Accelerated Surface Area & Porosimetry System (ASAP) 2020 by Micromeritics Instruments Inc. The glass analytical tube was filled with a dry sample of 50–100 mg. Prior to analysis, the samples were activated for 6 h at 120 °C in an argon gas environment, as determined by a thermogravimetric activation screening method (Fig. S15). The samples were then subjected to two stages of activation under vacuum (10⁻⁶mbar)—first to 60 °C for 2 h, followed by heating to 150 °C for 15 h, while maintaining an outgassing rate of <10 µbar per minute. Subsequently, the sample was backfilled with N₂ and transferred to the analysis port, where it was evacuated for at least an additional 120 min prior to commencing the pressure swing analysis, where the pressure is varied from 0 to 1 bar and each data point is taken after the samples have reached equilibrium. The Freundlich isotherm model and the Toth isotherm model (Fig. S8) were studied to gain insights into the adsorption results of CO₂ and N₂ on MEEG-CS aerogel.

Ideal Adsorption Solution Theory (IAST) has been employed in this study to assess multi-component (CO₂-N₂) adsorption systems using solely pure component experimental data⁶². The post-combustion process of power plants typically results in flue gas consisting of 75–85% N₂ and 15–25% CO₂. In this study, the partial pressures of CO₂ and N₂ are taken as 0.15 and 0.85, respectively, to determine the selectivity of CO₂ over N₂ and investigate whether the adsorption performance of MEEG-CS aligns with that of flue gas in humid power plants. The general procedure of calculating selectivity by the IAST model and the expression is given as⁶³:

$$s=\frac{q_i/q_j}{p_i/p_j}$$ 3

where S is the selectivity coefficient; q_i and q_j are amounts adsorbed of components i and j, respectively; p_i and p_j are partial pressure of components i and j, herein are 0.15 and 0.85, respectively.

Thermal regeneration via electrical heating

Joule heating was employed to thermally regenerate aerogels vis electrical heating, post CO₂ adsorption of the electrically conducting aerogels. The setup used was an Electro-Rheological Device (ERD) in the MCR 702 rheometer to regenerate the MEEG-CS aerogels, which allowed applying a variable electric current of up to 20 A. To control the regeneration temperature, we combined the 25 mm parallel plate measuring system with a Peltier device. Adsorption was performed at 25 °C for 30 min, and desorption was achieved by passing a DC current of 1 A through the aerogel, raising its temperature to 150 °C in around 20 s, and holding it at that temperature for 50 min. We performed 20 adsorption-regeneration cycles using the electrical heating regeneration method and measured the change in mass by weighing the aerogels before and after each cycle.

Author contributions

The manuscript was written through the contributions of all authors. All authors have approved the final version of the manuscript.

Peer review

Peer review information

Communications Materials thanks Dechao Wang and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editors: Natalia Shustova and Jet-Sing Lee. [A peer review file is available].

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s43246-025-00861-9.