Monolithic binder-free CO2 adsorbents with high adsorption capacity, selectivity, adsorption-desorption kinetics, and regenerability are highly desired to both reduce the environmental impact of anthropogenic CO2 emissions and purify valuable gases from CO2.
Microporous and Mesoporous Materials 323 (2021) 111236 Contents lists available at ScienceDirect Microporous and Mesoporous Materials journal homepage: www.elsevier.com/locate/micromeso Monolithic carbon aerogels from bioresources and their application for CO2 adsorption Shiyu Geng a, *, Alexis Maennlein a, Liang Yu b, Jonas Hedlund b, Kristiina Oksman a, c a Division of Materials Science, Department of Engineering Sciences and Mathematics, Luleå University of Technology, SE-971 87, Luleå, Sweden Chemical Technology, Department of Civil, Environmental and Natural Resources Engineering, Luleå University of Technology, SE-97 187, Luleå, Sweden c Mechanical & Industrial Engineering (MIE), University of Toronto, Toronto, ON, M5S 3G8, Canada b A R T I C L E I N F O A B S T R A C T Keywords: Monolithic carbon aerogels CO2 adsorbents Bioresources CO2/N2 selectivity Lignin Monolithic binder-free CO2 adsorbents with high adsorption capacity, selectivity, adsorption-desorption kinetics, and regenerability are highly desired to both reduce the environmental impact of anthropogenic CO2 emissions and purify valuable gases from CO2 Herein, we report a strategy to prepare monolithic carbonaceous CO2 ad sorbents from low-cost and underutilized bioresources, which enabled the formation of a delicate anisotropic, hierarchical porous structure With optimized material composition and processing conditions, the biobased carbon adsorbent demonstrated a CO2 adsorption capacity of 4.49 mmol g-1 at 298 K and 100 kPa, relatively weak adsorbent-adsorbate affinity, good CO2/N2 selectivity, and advantageous hydrophobicity against water vapor Moreover, the unique anisotropic porous structure provided high stiffness and good flexibility to the adsorbent in the axial and radial directions, respectively We confirmed that this type of carbon adsorbent could be packed in a column for dynamic CO2 capture independent of any binders, indicating its promising future for further development toward widespread utilization Introduction The capture and recovery of CO2 are vital tasks as the rate of anthropogenic CO2 emissions has overtaken the natural carbon cycle, posing risks to the climate [1] The separation of CO2 from other valu able gases has also attracted attention owing to the requirements for their purification For example, biogas contains 20%–50% of CO2 which needs to be removed to improve the calorific value and make biogas a competitive biofuel [2] Traditional CO2 capture methods are normally based on chemical absorption, such as amine scrubbing, which has high energy consumption owing to the need for solvent regeneration, suffers from amine degradation problems, and causes equipment corrosion [3, 4] By contrast, physical capture methods using porous adsorbents have been proposed and investigated since the 1980s [5–8], because they rely on molecular sieving effects and weak adsorbent-adsorbate interactions that lead to easier regeneration and a longer lifespan of the adsorbents Various materials such as zeolites, activated carbons, mesoporous sil icas, and metal-organic frameworks have been developed for CO2 adsorption and separation [9–15] Among these materials, activated carbons have several advantages, including low cost, large specific surface area, and high thermal and chemical stability [16,17] However, the use of conventional activated carbons for CO2 capture is limited by their relatively low adsorption capacity and large structural variation, depending on the base material [9] Other types of porous carbon materials have been investigated to improve the CO2 adsorption capacity of carbonaceous adsorbents, such as carbon molecular sieves [18,19], nitrogen-doped porous carbons [20–23], and carbon nanotubes [24–27] Although superb capacity and selectivity have been reported for many of these adsorbents, the use of non-renewable resources and complicated manufacturing processes still pose as significant obstacles for widespread applications Considering the sustainability requirements, high-performance carbonaceous ad sorbents produced from renewable bioresources via simple and straightforward processes are highly desired Recently, we have suc cessfully developed carbon aerogels from bio-based raw materials, lignin and cellulose nanofibers (CNFs), which showed great potential for use as CO2 adsorbents [28] Lignin as one of the main components of wood possesses a high aromatic content, and it is the byproduct of the pulp and paper industry most in need of utilization [29], and CNFs, which can be isolated from various plants, are promising green nano materials with large aspect ratios and strong mechanical properties [30] The derived monolithic carbon aerogels possessed a hierarchical porous * Corresponding author E-mail address: shiyu.geng@ltu.se (S Geng) https://doi.org/10.1016/j.micromeso.2021.111236 Received 17 March 2021; Received in revised form June 2021; Accepted June 2021 Available online 11 June 2021 1387-1811/© 2021 The Authors Published by Elsevier Inc This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/) S Geng et al Microporous and Mesoporous Materials 323 (2021) 111236 structure with homogeneous, tracheid-like macropores oriented in the same direction, which was attributed to the ice-templating and carbonization process that we used This interesting structure is ex pected to be favorable for CO2 capture because it contributes to the fast adsorption and desorption kinetics and also helps reach high adsorption capacity [9] Therefore, this type of carbonaceous adsorbent should be further modified and explored for higher CO2 capture efficiency The main aim of this study was to improve the CO2 adsorption per formance of lignin/CNF-based carbon aerogels to the next level, and to evaluate their properties as monolithic adsorbents The effects of different material compositions and processing factors, including the type of lignin, lignin/CNF ratio, solid content of the starting suspension, and carbonization time, on the final porous structure and properties of the carbon aerogels were systematically investigated The optimized adsorbent prepared without any activation process exhibited an excel lent CO2 adsorption capacity, which was competitive with many acti vated carbonaceous adsorbents synthesized using more complex procedures with lower yields reported in the literature [21,31–38] The CO2 selectivity against N2 and water vapor adsorption of the adsorbent were also measured, as they are important in gas separation and puri fication applications Moreover, as most of the conventional and previ ously investigated adsorbent materials are in powder form and their performance can be reduced by the binder blocking effect when applied with a binder, adsorbents in monolithic form are preferable [18] Consequently, the bulk behaviors of the lignin/CNF-based carbon aer ogels obtained from forward-step change breakthrough tests and me chanical tests are presented, which indicated that these monolithic carbon adsorbents derived from bioresources had the capability to be used without any additional materials or support in various CO2 capture applications Table Coding, composition, and carbonization time of samples prepared in this study Sample code Lignin type Composition (w/ w) Solid contenta (wt %) Carbonization timeb (h) Lignin TOCNF K75/253w-1h K85/153w-1h K85/156w-1h K85/153w-2h K85/153w-3h L85/153w-1h Kraft 75 25 Kraft 85 15 Kraft 85 15 Kraft 85 15 Kraft 85 15 3 Lignoboost 85 15 a Solid content of the related lignin/CNF suspensions without considering the weight of the added NaOH b Isothermal holding time at 1200 ◦ C newly developed LignoBoost process [39], named lignoboost lignin, was then chosen to investigate the effects of different types of lignin on the structure and properties of the derived carbon aerogels This is because the lignoboost lignin possesses higher purity and higher content of phenolic hydroxyl, enol ether and stilbene, and lower content of ash, methoxy group and β-O-4 structure compared to the traditional kraft lignin [40] The CNFs used in this study had an average diameter of 1.7 nm and a carboxylation degree of 0.72 mmol g− 1, which were well characterized in our previous work [28] The sample codes, material compositions, and processing parameters of all prepared carbon aero gels are presented in Table The whole carbonization process included three isothermal steps at 100, 500, and 1200 ◦ C in sequence, while the carbonization time mentioned in the following discussion is only related to the isothermal time at 1200 ◦ C Results and discussion 2.1 Synthesis of monolithic carbon aerogels The manufacturing process, from the lignin/CNF suspensions to the final carbon aerogels, is schematically shown in Fig 1a, and represen tative images of the lignin/CNF precursor and carbon aerogel are demonstrated in Fig 1b and c, respectively A commercial kraft lignin from traditional kraft pulping process was first used to determine the optimized lignin/CNF ratio, suspension solid content, and carbonization time, as shown in Table Another type of lignin generated from the 2.2 Physical and thermal properties of carbon aerogels and their raw materials The viscosity of the lignin/CNF suspensions, as shown in Fig 2a, varied significantly with the lignin/CNF ratio, solid content, and lignin type The K75/25-3w and K85/15-6w suspensions had viscosities of 96 and 91 mPa s, respectively, which were much higher than those of K85/ 15-3w (26 mPa s), due to the higher CNF content (25 wt% vs 15 wt% of dry weight) and solid content (6 wt% vs wt%) Compared to the K85/ 15-3w, the L85/15-3w suspension interestingly showed a lower viscos ity of 11 mPa s likely due to both the difference in the charge content between the lignoboost lignin and the kraft lignin and the different number of hydrophilic groups in them [41] It is obvious from Fig 1b and c that the lignin/CNF precursors contracted drastically during the carbonization, and the material yield and volume shrinkage were in the range of 33%–37% and 61%–68%, respectively, as illustrated in Fig 2b and Table S1 Among the carbon aerogels derived from the kraft lignin, K85/15-6w-1h had the highest yield (37%) and the lowest volume shrinkage (61%), caused by the higher solid content, while there was no considerable difference between the K85/15-3w carbon aerogels with various carbonization times at 1200 ◦ C (1, 2, and h) The carbon aerogel derived from lignoboost lignin (L85/15-3w-1h) also showed quite high yield and low volume shrinkage, especially when compared to K85/15-3w-1h, which had the same material composition and pro cessing condition except the lignin type Fig 2c reveals that the porosity of all carbon aerogels (calculated using Eq (1)) reached approximately 98%, indicating the super lightweight nature of the carbon aerogels, which is attributed to the ice-templating induced tracheid-like macropores Fig (a) Schematic preparation processing from lignin/CNF aqueous sus pensions to carbon aerogels, and representative illustrations of (b) a lignin/CNF precursor and (c) a carbon aerogel S Geng et al Microporous and Mesoporous Materials 323 (2021) 111236 Fig (a) Viscosity of the lignin/CNF suspensions (b) Yield and volume shrinkage of the carbon aerogels from their lignin/CNF precursors and (c) porosity of the carbon aerogels (solid columns: K85/15-3w samples with different carbonization time, columns with patterns: other samples) To individually understand the behaviors of the kraft lignin, the lignoboost lignin, and the CNFs during the carbonization process, their thermal degradation features were characterized by thermogravimetric analysis (TGA), as shown in Fig 3a and b Both types of lignin experi enced a small weight loss in the starting stage up to 100 ◦ C due to moisture evaporation, and their primary degradation occurred at 250–450 ◦ C caused by the release of volatiles including aromatics, al kyls, carbonyls, CO, and CO2 [42,43] It is obvious from the derivative thermogravimetry (DTG) curves that the kraft lignin reached the highest rate of weight loss at a lower temperature (310 ◦ C) than the lignoboost lignin (380 ◦ C), while the weight loss of the lignoboost lignin was much higher than that of the kraft lignin at the primary degradation stage The lignoboost lignin also showed better thermal stability at higher tem perature, as its weight loss rate gradually declined with increasing temperature and was close to after 730 ◦ C, while the weight of the kraft lignin continued to decrease This can be related to the higher yield of the lignoboost lignin-derived carbon aerogel after carbonization at 1200 ◦ C (Fig 2b) The possible reason is that the larger quantity of phenolic hydroxyl, enol ether, and stilbene structures present in the lignoboost lignin could trigger more crosslinking reactions at high temperatures compared to the kraft lignin [44,45], which resulted in a more stable carbon structure As a comparison, the CNFs started to degrade from 200 ◦ C, attributed to their thermally unstable anhydro-glucuronic acid units [46], which led to a lower residue content of 23% at 900 ◦ C than that of the kraft lignin (44%) and lignoboost lignin (35%) This indicates that the CNFs not only contributed to the shape stability of the lignin/CNF precursors, due to their fiber geometry with a large aspect ratio, but also acted as sacrificial templates for generating cavities during carbonization, which resulted in a hierarchical porous structure in the materials The thermal degradation behaviors of K85/15-3w and L85/15-3w precursors were also characterized to further compare the effects of the different types of lignin, and their TGA curves are exhibited in Fig S1 Due to the presence of the TOCNFs, both precursors had the primary degradation stage at a lower temperature region (200–400 ◦ C) with two main DTG peaks compared to that of the neat kraft lignin and lignoboost lignin (250–450 ◦ C) K85/15-3w pre cursor showed the highest rate of weight loss at a lower temperature than that of L85/15-3w precursor, which corresponds to the DTG curves of the neat lignin (Fig 3a) In addition, it can be noticed that the TGA residue content of L85/15-3w precursor at 900 ◦ C (32.2%) was even lower than the yield of the L85/15-3w-1h carbon aerogel carbonized at 1200 ◦ C (37.4%, Fig 2b) The possible reasons include that the isothermal step at 500 ◦ C during the carbonization process increased the thermal stability of the precursor owing to the lignin crosslinking effect, and some volatiles and tars generated during carbonization could not be removed as efficient as in the TGA test caused by the much bulkier sample size The carbon structure of the carbon aerogels was analyzed by Raman spectroscopy, and the results are shown in Fig 3c and Fig S2 All samples exhibited typical D (~1338 cm− 1) and G (~1584 cm− 1) bands of carbon materials, which are usually assigned to the breathing mode of disordered carbon atoms and the in-plane bond-stretching of carbon sp2 sites, respectively Farrari et al interpreted that in the case of amorphous carbon with nanocrystalline graphite, the intensity of D band is proportional to the graphitic cluster area, which is instead related to ordering [47] The inset of Fig 3c clearly demonstrates that both K85/15-3w-2h and K85/15-3w-3h exhibited higher D-band in tensity compared to K85/15-3w-1h, indicating that the size of graphite crystallites in the carbon aerogels was increased by extending the carbonization time to h, while it remained constant when the time was extended to h 2.3 Morphology and pore structure of carbon aerogels The morphology of both the cross section and the longitudinal sec tion (Fig S3) of the carbon aerogels was investigated using scanning electron microscopy (SEM), as illustrated in Fig All carbon aerogels Fig TGA curves of (a) kraft lignin and lignoboost lignin as well as (b) 2,2,6,6-tetramethylpiperidinyl-1-oxyl radical (TEMPO)-oxidized cellulose nanofibers (TOCNFs), and (c) Raman spectra of K85/15-3w carbon aerogels with different carbonization time (the intensity was normalized according to that of G bands) S Geng et al Microporous and Mesoporous Materials 323 (2021) 111236 Fig SEM images of (a) K75/15-3w-1h, K85/15-3w-1h and K85/15-6w-1h (up: cross section, bottom: longitudinal section; scale bar: 50 μm), (b) K85/15-3w carbon aerogels with different carbonization time (cross section; scale bar: 50 μm), and (c) carbon aerogels derived from different types of lignin (up: cross sec tion, bottom: longitudinal section; scale bar: 20 μm) exhibited an anisotropic structure with longitudinal macropores because of the ice-templating technique, and Figs S4 and S5 confirm that this unique structure remained homogenous at a relatively large scale (more than mm2 in cross section) Compared to K85/15-3w-1h, K75/25-3w1h, and K85/15-6w-1h showed macropores with a more regular shape and less wrinkled cell walls (Fig 4a), especially in the case of the former This can be associated with the higher CNF concentration and solid content in the starting lignin/CNF suspensions, which induced the for mation of a more rigid cell structure after ice-templating and carbon ization There was no distinct difference between the morphology of the carbon aerogels with different carbonization times, as shown in Fig 4b For the carbon aerogels with different types of lignin (Fig 4c), it is interesting that the aerogel with the lignoboost lignin (L85/15-3w-1h) also demonstrated less wrinkled and smoother cell walls, which could be attributed to the better thermal stability of the lignoboost lignin in the high-temperature range, corresponding to the above-mentioned yield/ volume shrinkage and TGA results Inspired by our earlier work [28], all carbon aerogels obtained in this study were washed with distilled water after carbonization to remove possible residual impurities and subsequently increase their porosity and surface area, which acted as key factors in CO2 adsorption To confirm the effects of washing, Brunauer− Emmett− Teller (BET) surface area analysis and energy-dispersive X-ray spectroscopy (SEM-EDX) were carried out for both unwashed and washed samples, and the results are shown in Fig 5, Table 2, and Table S2 The N2 adsorption isotherms for the BET analysis are shown in Figs S6 and S7 All washed samples demonstrated a much higher surface area compared to the unwashed ones (Fig 5), and one possible reason is the removal of salt compounds that expose the previously blocked pores This can be verified by the EDX data shown in Table 2, where the oxygen, sodium, and potassium contents in the kraft lignin-derived samples decreased considerably after washing, while the relative carbon contents increased However, the lignoboost lignin-derived carbon aerogel (L85/15-3w-1h) did not show a similar tendency, and the carbon content remained same, while the oxygen content was slightly increased by washing This indicates that removing salts was not the only reason of the increased surface area, but the removal of some tars generated during carbonization could also have contributed K85/15-3w-1h showed a higher surface area than that of K75/25-3w-1h and K85/15-6w-1h, owing to its more accessible micro pores (563 m2 g− of micropore-surface area after washing compared to 290 and 185 m2 g− 1, respectively; Table S2), which is probably caused by the different cell wall structure (Fig 4a), inducing different heat Fig BET surface area and pore size of all carbon aerogels before and after washing (solid columns: K85/15-3w samples with different carbonization time, columns with patterns: other samples) transfer and carbonization effects [48] Moreover, it was obvious that increasing carbonization time from h to h significantly reduced the surface area of the carbon aerogels and slightly increased the average pore size (from 1.6 to 1.9 nm, after washing) This was likely because more severe migration of carbon atoms resulted in pore merging in the samples with a longer carbonization time at 1200 ◦ C [49] 2.4 Adsorption properties of carbon aerogels The CO2 adsorption isotherms of all the washed carbon aerogels were measured in the 0–120 kPa pressure range and at three different tem peratures, which were 273, 298, and 323 K Fig 6a and b reveal that L85/15-3w-1h surprisingly exhibited a superior CO2 adsorption capacity compared to K85/15-3w-1h at all temperatures, which at 273 K reached 1.94 and 6.28 mmol g− at 10 and 100 kPa, respectively This was not in agreement with the BET surface area results (Fig 5), in which L85/153w-1h showed a drastically lower surface area (380 m2 g− compared to 643 m2 g− 1) This phenomenon implied that a large number of ultramicropores could be present in L85/15-3w-1h, which were accessible for CO2 while not detectable in BET measurements using N2 at 77 K because S Geng et al Microporous and Mesoporous Materials 323 (2021) 111236 Table Elemental composition (in at %) of the carbon aerogels before and after washing according to the SEM-EDX Sample coding K75/25-3w-1h K85/15-3w-1h K85/15-6w-1h K85/15-3w-2h K85/15-3w-3h L85/15-3w-1h Unwashed Washed C O Na Si S K C O Na Si S K 95.8 94.5 94.7 97.7 97.9 97.6 3.1 3.8 4.0 1.7 1.7 1.6 0.5 – 0.8 0.1 0.1 0.4 0.1 – – – – – 0.4 0.4 0.4 0.3 0.2 0.4 0.1 1.3 0.1 0.2 0.1 – 98.0 97.8 97.1 98.2 98.4 97.4 1.4 1.9 2.2 1.4 1.3 2.2 0.1 – 0.3 – – 0.1 0.1 – – – – – 0.4 0.2 0.4 0.3 0.3 0.3 – 0.1 – 0.1 – – Fig (a,b) CO2 adsorption isotherms with Langmuir model-fitting curves of K85/15-3w-1h and L85/15-3w-1h at 273, 298 and 323 K; and (c,d) the quantity of adsorbed CO2 at the pressures of 10 and 100 kPa of all prepared carbon aerogels at the various temperatures used (i: K85/15-6w-1h, ii: L85/15-3w-1h, ii: K85/15-3w2h, iv: K85/15-3w-3h, v: K75/25-3w-1h, vi: K85/15-3w-1h) of the smaller quadrupole moment and more diffusional restrictions of N2 when compared to CO2 [17,50,51] The data of CO2 adsorption ca pacity of all samples as a function of BET surface area are summarized in Fig 6c and d and Table S3 Among the kraft lignin-derived carbon aerogels, K85/15-3w-1h possessed the largest surface area and showed the highest CO2 adsorption capacity at 10 kPa, because of its large quantity of accessible micropores (Table S2) However, at 100 kPa, a similar CO2 adsorption capacity was observed for all kraft lignin-based samples, except for K85/15-6w-1h, which had a much smaller surface area To further investigate this, Fig S8 compares the CO2 adsorption isotherms recorded at 273 K of the K85/15-3w carbon aerogels with different carbonization times With increasing carbonization time, the isotherm slope in the relatively high-pressure range increased positively, and the adsorption quantity of K85/15-3w-3h exceeded that of both K85/15-3w-1h and K85/15-3w-2h at pressures higher than 85 kPa This was likely due to the larger pore size of the samples with longer carbonization times (Table S2), resulting in the condensation of CO2 molecules in the pores at high pressure A similar behavior, that of a steep isotherm slope at high pressure, was also observed for K75/25-3w-1h (Fig S9), which had a pore size of 2.1 nm In addition, by calculating the CO2 adsorption enthalpy (ΔH) and entropy (ΔS) of the samples from their isotherms at various temperatures (Eq (2) and Eq (3)), it can be seen in Fig that the absolute values of ΔH (|ΔH|) of K75/15-3w-1h, K85/15-3w-2h, and K85/15-3w-3h are greater than the other two kraft lignin-based samples, which illustrates that they can provide stronger affinity for CO2 L85/15-3w-1h exhibited a relatively low affinity for CO2 (|ΔH| is 19.98 kJ mol− 1), while still showed promising CO2 adsorption capacity, even when compared to many literature-reported carbonaceous adsorbents prepared with various activation processes (Table 3), revealing its excellence not only in CO2 capture but also in adsorbent regeneration As L85/15-3w-1h outperformed all the other samples, showing the highest CO2 adsorption capacity, it was selected to evaluate the CO2/N2 selectivity and water vapor adsorption, as well as the monolithic prop erties, including the column breakthrough adsorption capacity and mechanical properties Fig 8a shows the comparison between the CO2 and N2 adsorption isotherms of L85/15-3w-1h at 298 K The CO2/N2 selectivity was estimated to be 21 from the ratio of the slope of the S Geng et al Microporous and Mesoporous Materials 323 (2021) 111236 the typical water vapor adsorption behavior of hydrophobic carbon materials with a hierarchical porous structure [56,57] The adsorption capacity of L85/15-3w-1h was recorded as 0.46 mmol g− at a pressure of kPa, which is considerably lower than that of bituminous coal-based activated carbons at the same temperature (approximately 1–2 mmol g− at 298 K) [58,59] This is probably because the high carbonization temperature (1200 ◦ C) used for the carbon aerogels led to a more hy drophobic surface with fewer sorption sites for water molecules Fig 8c illustrates the results from the forward-step change break through experiment for the L85/15-3w-1h with a 10 kPa CO2/90 kPa N2 mixture, which was used to simulate the realistic conditions of postcombustion capture The test setup is depicted in Fig S11, where several carbon aerogels were loaded in the column to achieve a suffi cient ratio of adsorbent length to diameter By integrating the area be tween the sample curve and the curve of the empty column, a dynamic CO2 adsorption capacity of 0.67 mmol g− was obtained The value was lower than that obtained from the equilibrium measurement at the same pressure and temperature (1.11 mmol g− at 10 kPa CO2 and 298 K) because of the competitive adsorption from N2 According to the abovementioned CO2/N2 selectivity of L85/15-3w-1h (a value of 21), its theoretical CO2 adsorption capacity under a 10 kPa CO2/90 kPa N2 at mosphere can be calculated as 0.63 mmol g− 1, which corresponds very well with the measured result Fig Enthalpy (ΔH) and entropy (ΔS) of CO2 adsorption of all carbon aerogels (solid columns: K85/15-3w samples with different carbonization time, columns with patterns: other samples) Table Comparison of CO2 adsorption capacity at 100 kPa and 298 K between L85/153w-1h and other carbonaceous adsorbents made from various precursors pre pared by different processes, as reported in literature 2.5 Mechanical properties of monolithic carbon aerogels Material code Carbon resource Processing CO2 adsorption capacity (mmol g− 1) Reference CP-2-600 Polypyrrole 3.84 [21] MFB-600 2.25 [31] GKOSA50 Melamine and formaldehyde Olive stones Nitrogen-doping and chemical activation Pyrolysis 2.43 [32] AA750 Almond shell 2.66 [33] HCMDAH-1900-1 Resorcinol and formaldehyde 3.30 [34] Conclusions AS-2-600 Sawdust NCP-800 Coal tar pitch Y–K-600 Yeast L85/153w-1h Lignoboost lignin and nanocellulose Pyrolysis and physical activation Pyrolysis and physical activation Nitrogen-doping, pyrolysis, and physical activation Hydrothermal carbonization and chemical activation Nitration and pyrolysis Pyrolysis and chemical activation Pyrolysis The mechanical properties of L85/15-3w-1h in both the axial and radial directions were characterized by compression testing, and the acquired stress-strain curves are shown in Fig The sample exhibited distinct anisotropic mechanical behaviors, which was attributed to the ice-templating induced tracheid-like porous structure The elastic modulus of the sample in the axial direction was as high as 2.64 MPa (Table S4) with a specific elastic modulus of 61.2 kNm kg− (calculated using Eq (4)), while in radial direction the sample showed good flexi bility, reaching 50% of the strain before collapse The excellent me chanical properties of the carbon aerogels with a porosity as high as 98% make them a viable choice for CO2 capture applications without any supporting materials or binders, demonstrating their potential as inde pendent CO2 adsorbents in the future 4.82 [35] 2.20 [36] 4.77 [37] 4.49 This work In summary, we have demonstrated that carbon aerogels based on lignin and CNFs with anisotropic, hierarchical porous structures were successfully prepared via a straightforward procedure combining icetemplating and carbonization By tailoring the CNF concentration, solid content, and carbonization time, the structure of the carbon aer ogels, including the carbon structure, morphology, porous structure, and surface area, were varied, which consequently led to different CO2 adsorption capabilities The different lignin structure generated from different types of lignin affected their thermal degradation behaviors during carbonization, also resulting in significant variations in the structure and properties of the obtained carbon aerogels The carbon aerogel containing lignoboost lignin reached a CO2 adsorption capacity of 4.49 mmol g− at 298 K and 100 kPa, which is competitive among previously reported carbon-based adsorbents, while possessing a good CO2/N2 selectivity of 21 and a low water vapor adsorption capacity up to 1.5 kPa of vapor pressure Furthermore, we have also discovered that carbon aerogels had excellent mechanical properties and could be used alone in a column for CO2 capture without the need for any additional binders, which is expected to initiate the further development of monolithic carbonaceous CO2 adsorbents with large adsorption capac ity, high selectivity, fast adsorption-desorption kinetics, and easy regeneration isotherm linear regions in the low pressure range (≤10 kPa), which is higher than or comparable to that of many modified carbon materials [22,52,53] The water vapor adsorption of the sample was evaluated at 298 K in the 0–3.169 kPa pressure range (up to the saturated vapor pressure) [54] The isotherm is shown in Fig 8b and can be described as a Type-V isotherm according to the IUPAC classification [55] This in dicates that the amount of adsorbed water was relatively low at pres sures below 1.5 kPa, while significant capillary condensation of water molecules occurred at higher pressures Finally, the sample reached a high-water adsorption capacity due to pore filling This corresponds to S Geng et al Microporous and Mesoporous Materials 323 (2021) 111236 Fig (a) CO2 and N2 adsorption isotherms of L85/15-3w-1h at 298 K, and (b) water vapor adsorption isotherm and (c) forward-step change breakthrough curves with CO2/N2 mixture (volume ratio of 10/90) of L85/15-3w-1h and the empty column at 298 K 4.3 Preparation of carbon aerogels The carbon aerogels were prepared by the carbonization of the lignin/CNF precursors under a nitrogen atmosphere using a tube furnace (RHTC-230/15, Nabertherm GmbH, Lilienthal, Germany) The heating procedure included temperature ramp from room temperature to 100 ◦ C with an isothermal holding time of h, ramp from 100 to 500 ◦ C and holding for 100 min, and a final ramp up to 1200 ◦ C with different holding times of 1, 2, and h The heating rate of all ramp steps was ◦ C/min Then, the carbonized samples were washed with distilled water five times at 30-min intervals and dried in an oven at 80 ◦ C overnight to obtain the final carbon aerogels 4.4 Characterizations The viscosity of the lignin/CNF suspensions was measured using an SV-10 Vibro viscometer (A&D Company, Tokyo, Japan) at 22 ◦ C TGA was performed using a TA Q500 thermogravimetric analyzer (TA In struments, New Castle, DE, USA) under a nitrogen atmosphere, with a temperature range from room temperature to 900 ◦ C and a heating rate of 10 ◦ C/min The porosity of the carbon aerogels was calculated as follows: ( ) ρ∗ P= 1− × 100% Eq Fig Representative stress-strain curves of L85/15-3w-1h from compression testing in both axial and radial directions Material and methods 4.1 Materials ρ Kraft lignin with a low sulfonate content (Mw of ~10,000) was purchased from Sigma-Aldrich, Sweden AB Lignoboost lignin was supplied by Domtar Plymouth pulp mill (NC, USA) with a Mw of 6772 and a purity of 96.5% [40] Cellulose nanofibers (CNFs) were prepared through 2,2,6,6-tetramethylpiperidinyl-1-oxyl radical (TEMPO)-me diated oxidation treatment followed by a homogenization process, as described in our previous study [28] Sodium hydroxide (NaOH, pure pellets) was purchased from Merck KGaA, Germany All chemicals were used as received where ρ and ρ* denote the density of the solid carbon (2.1 g cm− 3) [54] and the bulk density of the carbon aerogels, respectively The carbon structure was characterized by Raman spectroscopy using a Bruker Senterra dispersive Raman spectroscope (Bruker Corp., Billerica, MA, USA) with a 533 nm laser beam SEM was conducted using a JEOL JSM 6460 L V scanning electron microscope (JEOL Ltd., Tokyo, Japan) The elemental composition of the samples was examined by SEM-EDX equipped with a silicon drift detector (Oxford X-MaxN 50 mm2, Ox ford Instruments, Oxfordshire, UK) BET surface area analysis was car ried out according to N2 adsorption tests at 77 K using a Gemini VII 2390a analyzer (Micromeritics Instrument Corp., Norcross, GA, USA) The samples were degassed at 300 ◦ C for h prior to the tests CO2 adsorption measurements were performed with an ASAP 2020 Plus BET analyzer (Micromeritics Instrument Corp., Norcross, GA, USA) The pressure ramp was from to 120 kPa at 273, 298, and 323 K, respec tively, after the degas step Based on the adsorption isotherm data at low pressure (