Evaluation of amine-based solid adsorbents for direct air capture: a critical review

Abstract

Direct air capture (DAC) emerges as a new technology that can contribute to “negative carbon emission”. Recent progress in surface chemistry and material synthesis has allowed a new generation of CO₂ adsorbents that can drive the future of DAC and its wide-ranging deployment. This review is intended to shed light on the recent developments in porous amine-based solid sorbents for direct air CO₂ capture, adsorbent preparation and characterization, CO₂ capture under dry and humid conditions, CO₂ adsorption kinetics, adsorption thermodynamics, sorbent regeneration, cyclic stability, essential regeneration techniques, and techno-economic analysis for CO₂ capture from air.

---

Debashis Panda

Debashis Panda obtained his Ph.D. in Mechanical Engineering from the Indian Institute of Technology Indore, India under the joint supervision of Prof. Sanjay Kumar Singh and Prof. E. Anil Kumar. His research mainly focuses on the synthesis and characterization of solid adsorbents for CO2 capture from various sources including flue gas, biogas, and ambient air. He has also focused on the scalability of adsorbent and explored suitable separation techniques including thermal swing adsorption for direct air CO2 capture applications. His research activity is highly interdisciplinary lying between the areas of engineering and chemistry.

Vaishnavi Kulkarni

Vaishnavi Kulkarni received her M.Sc. in Chemistry in 2017 from Savitribai Phule Pune University, India. Currently, she is a Ph.D. student in the Department of Chemistry, Indian Institute of Technology Indore, India under the supervision of Prof. Sanjay Kumar Singh. Her research expertise involves the synthesis of mesoporous/microporous as well as hierarchical materials for CO2 capture. She is currently working on the synthesis of hierarchical silica, metal–organic frameworks, and their further modification for direct air CO2 capture applications.

Sanjay Kumar Singh

Sanjay Kumar Singh obtained his PhD in Chemistry from A.P.S. University, Rewa, India. He had consecutive postdoctoral stays at AIST, Osaka, Japan (as a JSPS postdoctoral fellow, and later as AIST postdoctoral scientist with Prof. Qiang Xu) and at Karlsruhe Institute of Technology, Karlsruhe, Germany (as Alexander von Humboldt postdoctoral fellow with Prof. Peter W. Roesky). In 2012, he joined the Indian Institute of Technology Indore, India, as an Assistant Professor in Chemistry, and since Feb. 2022 he is a Professor at the same institution. His research focuses on the design and development of new and efficient catalysts and materials for CO2 capture and utilization, hydrogen production and storage, and biomass transformation to value-added chemicals and fuels.

---

1 Introduction

The pursuit of socio-economic growth for the development of human civilization has imparted huge dependence on conventional fossil fuels. This has led to massive greenhouse gas (GHG) emissions such as CO₂, causing global climate change.¹ Besides the combustion of fossil fuels, a considerable amount of CO₂ is also generated during various processes such as cement production, bio/natural gas sweetening, and electricity generation from geothermal vents.² As a result, the atmospheric CO₂ concentration has risen dramatically from 278 ppm at the start of the Industrial Revolution to >400 ppm today.^3,4 In fact more than 2000 Gt CO₂ has been emitted by humans since the Industrial Revolution.⁵ This has increased the earth's average temperature by 0.6 to 1 °C. Without any intermediate strategy for negative carbon emissions, the current global GHG emission is expected to increase by about 11% by 2030, compared to 2010 levels.⁶ The global mean temperature has risen by 1 °C after the pre-industrialization period.⁷ Therefore, 10 Gt CO₂ per year must be removed from the atmosphere each year until 2050, and then 20 Gt CO₂ per year from 2050 to 2100 to keep warming below 1.5 °C (Fig. 1).⁸ In this context, various carbon capture storage (CCS) technologies such as forestation, enhanced weathering, bioenergy and carbon capture and storage (BECCS), pre and post-combustion CO₂ capture, and direct air CO₂ capture have been deployed on a large scale, which could address both carbon emission problems as well as generation of renewable fuel from captured CO₂ to fulfill energy demands.^9–11

Fig. 1 Atmospheric CO₂ concentration and CO₂ mitigation timeline. Blue shading represents the amount of negative emissions likely needed in this scenario. Reproduced with permission.¹⁰ Copyright 2019, Frontiers in Climate.

DAC is recently introduced as modern CCS, a complementary approach to capture CO₂ directly from the air at a low energy penalty and convert it into CO₂-neutral liquid or gaseous hydrocarbon fuels.¹² DAC is geographically agnostic (i.e. located anywhere) and essential for achieving net-zero and thus is now accepted by the world leaders meet at Glasgow (for COP 26).⁵ Planting trees and rewilding are not enough to reduce the CO₂ content in the air up to a permissible limit. Instead, developing cost-effective DAC technology and reducing fossil fuel usage may be a solution to achieve net-zero.¹³ Nearly 40 direct air capture plants established by companies such as Climeworks and Carbon Engineering are in service around the world, capturing over 9000 tons of CO₂ per year directly from the air using various technologies.^14,15 To have a meaningful impact of carbon removal on climate, DAC needs to be economical and practical to capture CO₂ at gigatons level. With a carbon capture capacity of 1 Mt CO₂ per year, Ozkan et al. estimated that 1250 DAC plants would be required to remove 25 Gt of CO₂ by 2030.⁵ In this context, selecting a suitable CO₂ capture technology for DAC is tricky, as there is always a trade-off between performance and capture cost. The most important parameter is the low concentration of CO₂ in the air (0.04%) compared to point sources (such as flue gas ∼10–15%). Therefore, the kinetics and thermodynamics of gas separation from the air are less favorable.¹⁵ In general, both absorption and adsorption processes are widely used in DAC.¹⁶ Adsorption processes, using solid sorbents capable of capturing CO₂ from any stationary sources or atmosphere have shown many potential advantages,¹⁶ compared to other conventional CO₂ capture techniques using aqueous amine solvents, such as CO₂ scrubbing using aqueous N-methyl ethanolamine (MEA) solution.¹⁷ The high heat capacities of these solutions and the solvent losses due to evaporation^18,19 make the regeneration process highly energy-intensive and expensive. Significantly, the CO₂ scrubbing solutions such as the aqueous MEA solution, which has been used in chemical industries for several years, might not be the best choice for CO₂ absorption from the air, because of a large volume of gas (about ∼300 times)²⁰ to be treated and the subsequent loss of water and low-boiling MEA from the system. A recent study claimed that to capture 1 ton of CO₂ from the air using a conventional monoethanolamine (MEA) solution, the overall cost of $1690 would be required.²¹ In this context, various studies were carried out to reduce the regeneration energy during the desorption of CO₂ from MEA solution. For instance, Zhang et al.²² synthesized three distinct metal oxide-modified KIT-6 catalysts to catalyze the regeneration process of CO₂-loaded MEA solution. The group found that KIT-6-based catalysts can reduce the relative regeneration energy consumption by 20.2–33.4% while increasing the CO₂ desorption capacity by 31.0–49.3%. In line with this, Hu et al.²³ reported that the trio-amine blends MEA-1-dimethylamino-2-propanol/piperazine (1DMA2P–PZ) with different 1DMA2P/PZ mole ratios were formulated and applied for the post-combustion CO₂ capture. In addition, the blend-3 containing (MEA : 1DMA2P : PZ = 3 M : 2.5 M : 0.5 M) demonstrated outstanding performance in terms of relative heat duty and reduced the energy penalty by 29.4–55.4% when compared to the 5 M MEA solution. Chen et al.²⁴ investigated the absorption and desorption performance of three different aqueous MEA/MDEA/AMP (2-aminoethanol/N-methyldiethanolamine/2-amino-2-methyl-1-propanol) blends, and blend-2 (5 M, 1/2/2) showed a desorption rate up to 5 times higher, a 58.5% lower heat duty and a 57.6% lower relative cost compared to 5 M aqueous MEA. Similarly, Zhang et al.²⁵ reported the CO₂ absorption heat of tertiary amine 2-(dimethylamino)-2-methyl-1-propanol (2DMA2M1P), found to be 36.3 kJ mol⁻¹, which was quite lower than MEA and thus can be an efficient amine solvent for CO₂ capture. Despite significant progress in MEA-based CO₂ capture, liquid-based hydroxide solution (NaOH or KOH) is the only current commercial absorption technology capable of capturing CO₂ from the atmosphere to form water and carbonate. Later, carbonates are treated with lime (Ca(OH)₂) to precipitate CaCO₃, which is then heated at about 900 °C to release a high purity CO₂ gas for further sequestration and utilization.²⁶ However, the footprint of a liquid-based hydroxide solution adsorption system is quite bulky, and the regeneration process is also highly energy-intensive and expensive.^27,28 According to the American Physical Society (APS), the capture cost for a realistic design employing liquid hydroxide solution DAC is roughly 600 $ per tCO₂.²⁹

In this context, amine-based solid adsorbents are considered to be technically suitable and feasible for DAC due to their ability to capture CO₂ at low concentrations and even in the presence of moisture.^30–32 The improved performance characteristics such as easy handling, applicability to microgravity environments, and regenerability in either pressure swing (PSA) or temperature swing (TSA) modes expanded the application of amine-based adsorbents for CO₂ capture ranging from alkali fuel cell (AFC) to space life support applications.^33,34 Notably, such DAC processes with multiple pilot plants are already being developed and commercialized by two established startups: Climeworks and Global Thermostat.³⁵ Since 2006, there has been a significant increase in the number of publications in the area of DAC (Fig. 2), focusing mainly on the development of various porous materials and energy-efficient separation techniques for CO₂ capture from air.

Fig. 2 Number of journal articles published from 2006 to 2022 in the area of DAC (data collected from Scopus using the keywords for search: ‘DAC’, ‘direct air capture’, ‘direct air CO₂ capture’ and ‘CO₂ capture from air’).

A general representation of amine-based adsorbents for direct air CO₂ capture is illustrated in Fig. 3, where the amine-based adsorbents capture CO₂ from the air at ambient conditions. Subsequently, after CO₂ saturation, the sorbents are evacuated at a pressure in the range of 20–400 mbar and heated at a low temperature of 80–130 °C to desorb CO₂.^36,37 Further, the cost of CO₂ capture using amine-based sorbents would significantly reduce by regenerating the sorbent by utilizing waste heat or renewable energy sources. For instance, Fasihi et al.³⁵ estimated that the cost of a low-temperature TSA-based DAC (LT DAC) process using amine-based sorbent would reduce to 133 € per tCO₂ in 2020, 60 € per tCO₂ in 2030, 40 € per tCO₂ in 2040 and 32 € per tCO₂ in 2050, if powered by hybrid PV–wind–battery systems for Moroccan conditions or using waste heat.

Fig. 3 Schematic representation of amine-based adsorbent for direct air CO₂ capture.

Amine-based solid adsorbents, consisting of a mixture of amine and porous solid support, exhibit better CO₂ capture performance due to the mutual benefit of CO₂ chemisorption by amines and CO₂ physisorption by the remaining solid support. In this regard, physisorbents such as zeolites,³⁸ silica,^39–41 and MOFs,^42,43 are typically used as solid support to produce amine-based solid adsorbents. The interaction mechanism of CO₂ with amine (especially MEA) is well explored in the literature,^44–46 for instance by Caplow et al.⁴⁷ and Danckwerts et al.⁴⁸ Typically, reaction mechanism for CO₂ adsorption over amine based adsorbents can be categorized into three types (as shown in eqn (1)–(6)).^44–46

Zwitterion mechanism (type I): in the zwitterion mechanism, amine reacts with CO₂ to form a zwitterion (eqn (1)) in the first step,^47–49 and then the zwitterion is deprotonated by a free base (such as amine, OH⁻, or H₂O) to form carbamate (in dry condition),⁵⁰ and/or bicarbonate (under humid condition).⁵¹

CO₂ + RNH₂ ⇌ RNH₂⁺COO⁻ (E_a = ∼40–50 kcal mol⁻¹) (1)

RNH₂⁺COO⁻ + RNH₂ ⇌ RNHCOO⁻ + RNH₃⁺ (E_a = ∼13.2 kcal mol⁻¹) (anhydrous condition) (2)

CO₂ + RNH₂ + H₂O ⇌ RNH₃⁺ + HCO₃⁻ (E_a = ∼21 kcal mol⁻¹) (hydrous condition) (3)

Single-step mechanism (type II): in the single-step termolecular mechanism, proposed by Da Silva and Svendsen,⁵² amine reacts with CO₂ and a base simultaneously without the formation of any stable isolable zwitterion species. In this mechanism, instead of amine, water molecules act as a proton acceptors.^49,53

CO₂ + RNH₂⋯B ⇌ RNHCOO⁻⋯BH⁺ (4)

Carbamic acid mechanism (type III): carbamic acid mechanism also follows a two-step reaction pathway, wherein, initially amine reacts with CO₂ to form carbamic acid, and then carbamic acid reacts (proton transfer) with another amine to form carbamate.^54,55

CO₂ + RNH₂ ⇌ RNHCOOH (5)

RNHCOOH + RNH₂ ⇌ RNHCOO⁻ + RNH₃⁺ (6)

Among these three mechanisms, the zwitterion mechanism is mostly referred to in research articles⁴⁴ while addressing the CO₂ capture by amines whether in solutions or solid adsorbents under dry or humid conditions.⁵¹ Similar to liquid amine, species such as carbamic acid, alkylammonium carbamate, and/or bicarbonate are also identified during CO₂ adsorption over amine-based adsorbent, depending upon the amine loading and adsorption conditions⁵⁶ Kortunov et al.⁵⁷ reported that over the adsorbent surface, the nucleophilic attack of amine (Lewis base) on CO₂ carbon resulted in the formation of zwitterion RH₂N⁺–COO⁻ (referred as 1,3-zwitterion, where 1 and 3 indicate the positive and negative centers, respectively). Further, according to the working conditions, 1,3-zwitterion may transform to ammonium carbamate via intermolecular proton transfer with amine (Brønsted base) or afford carbamic acid via intramolecular hydrogen transfer. Notably, under anhydrous conditions, two primary amines reacts with CO₂ to yield ammonium carbamate.⁵⁸ For the reaction takes place in the presence of water; water acts as a nucleophile, reacting with electrophilic CO₂ to form carbonic acid^57,59 and which further produce carbonate/bicarbonate by reacting with amine.^57,60 In addition, water can hydrate carbamate anion to form carbonate or bicarbonate by releasing the amine molecule for further reaction with CO₂.^57,61 Hence, the maximum theoretical amine efficiency (molar CO₂/N ratio) is 0.5 for anhydrous and 1.0 for hydrous conditions.⁶² Primary and secondary amines can react with CO₂ in both dry and humid conditions, whereas, tertiary amine reacts with CO₂ only under humid conditions.⁶³ Further, the formation of carbamic acid or carbamate depends upon amine loading (or amine surface density),^58,64 when there is a large amine loading on the substrate, alkylammonium carbamate predominates,^64–67 while carbamic acid or hydronium carbamate is favored under low amine loading.^58,64,66 Under the humid condition, water can also assist the nucleophilic attack of CO₂ with the amine to generate hydronium carbamate, or it can act as a nucleophile that amines assist in producing ammonium bicarbonate at expense of carbamate.⁵¹ For highly hindered amine or tertiary amine, water becomes a more competitive Lewis base, resulting in the production of ammonium bicarbonate in comparison to carbamate.⁵¹

Various probing techniques such as Fourier transform infrared (FTIR),^{65,66,68–7013}C solid-state NMR,^{64,67,71–7315}N NMR,^74,75 and X-ray diffraction⁷⁶ are used to characterize the species form during CO₂ adsorption on amine-based adsorbents. However, the mechanisms of CO₂ interaction with amines, present in solution form or inside the adsorbent are still unclear. For instance, the zwitterion intermediates have been repeatedly reported in literature but experimentally not detected.^45,51 Various researchers have tried to elucidate the amine–CO₂ plausible reaction mechanism in terms of the energy barrier of 1,3-zwitterions. The activation energy barrier comes in the order of 40–50 kcal mol⁻¹ for a 1 : 1 reaction between CO₂ and amines.^54,66,69 For instance Said et al.⁵¹ reported the energy barrier (Gibbs activation energy resp.) E_a (Δ_rG* resp.) of 1,3-zwitterions to be 42.8 (44.8 resp.), and 40.0 (41.0 resp.) kcal mol⁻¹ for CO₂ interaction with primary and secondary amine (gaseous stage) respectively through a four-membered mechanism (Fig. 4). Activation barrier reduces further for the interaction of CO₂ and amine diluted with a suitable solvent such as CCl₄ (41.8 versus 44.8 kcal mol⁻¹ in the gas state) or water (31.3 kcal mol⁻¹). On the other hand, when CO₂ reacts with a diamine such as N-methylmethanamine in the gaseous state, catalytic assistance of a Brønsted base facilitates the reaction by decreasing the energy barrier (26.5 kcal mol⁻¹). Similarly, activation barrier of 21.0 kcal mol⁻¹ is associated with the formation of bicarbonate and carbonate by the reaction of CO₂ and water in the presence of aqueous amine. Hence, Said et al.⁵¹ proposed that the formation of 1,3-zwitterions may not prefer a four-member mechanism due to the high activation barrier (>30 kcal mol⁻¹) rather may follow a unified six-membered mechanism. The six-membered mechanism involves the nucleophilic attack of CO₂ by a Lewis base (amines or water) with hydrogen transfer or exchange assisted by a Brønsted base (amines or water). The plausible CO₂–amine interaction mechanism and associated activation energy, and possible reaction mechanism for CO₂–amine–water interactions are shown in Fig. 4.⁵¹

Fig. 4 (a) Illustrative CO₂–amine interaction pathways, involving (b) four-membered and (c) six-membered transition states. Reproduced with permission.⁵¹ Copyright 2020 American Chemical Society.

Although CO₂ adsorption on amine-based adsorbents is a well-established technique with extensive research, particularly on capturing high-concentration CO₂ from point sources such as flue gas and the cement industry, the material design for the DAC application, where CO₂ concentration is very low (0.04%), is still trivial. One needs to be more careful while choosing the best adsorbent and the evaluation parameters which will basically decide the overall CO₂ capture performance. Readers can find some interesting reviews on material development and techno-economic analysis of multiple techniques used for direct air capture.^{15,16,26,32,77–80} However, only limited focused work appeared on exploring the evaluation parameters for amine-based adsorbents in DAC applications.¹⁶ This review addresses the recent developments in amine-based porous solid sorbents for CO₂ capture, adsorbent preparation, and characterization, evaluation parameters such as CO₂ capture under dry and humid conditions, CO₂ adsorption kinetics, adsorption thermodynamics, sorbent regeneration, cyclic stability, essential techniques and economic analysis for CO₂ capture from air.

2 Synthesis of amine-based solid adsorbent for direct air CO₂ capture

DAC via amine-based adsorbent has gained significant interest nowadays due to their high CO₂ adsorption performance at ultra-dilute CO₂ concentrations and even in the presence of moisture.^{30,39,41,81,82} However, the performance of amine-based adsorbent predominantly depends upon its synthetic procedure and process parameters. The amine-based adsorbents used for DAC are typically synthesized in four ways: (i) physical impregnation (class I adsorbents), (ii) chemical grafting (class II adsorbents), (iii) in situ polymerization (class III adsorbents) and (iv) combination of both chemical grafting and physical impregnation (class IV adsorbents)¹⁶ as displayed in Fig. 5.

Fig. 5 Schematic representation of amine-based porous solid adsorbent. Adapted with permission from ref. 100. Copyright 2015 American Chemical Society.

Typically, in the physical impregnation method, the selected amine is first added to a solvent (such as water or any organic solvents e.g., methanol or ethanol) and uniformly dispersed by mechanical or magnetic stirring. Later, the desired quantity of porous solid support is added to the amine solution, followed by continuous stirring (for a specific duration) to disperse amine into the pores of the support. The resultant solution is finally filtered and the solvent is removed under pressure to obtain the amine-based solid adsorbent, where amines are expected not to have any direct chemical bonding with the support.⁸³ The loading of amine inside the pores of the support depends upon the available pore volume of the support and the amine density. The amine viscosity greatly influences its dispersion in the porous support system and the resulting adsorbent's degree of saturation. Further, the pore size and morphology restrict the molecular and kinetic diffusion of an amine into the pores of the support.^39,84 Mesopore and macropore volume (larger voids) are majorly responsible for containing the amine inside the sorbent pores/voids.³⁹ The voids present inside the solid support may contain amine; however, the residual amine species will agglomerate on the support surface once the percolation limit has been reached. The benefit of the physical impregnation method lies in its simple and facile synthesis process and the ability to accommodate large quantities of amine in the adsorbent having a high pore volume. However, low amine stability due to amine degradation and amine evaporation are the significant drawbacks that impede the adsorbent application in DAC. Amine-impregnated adsorbents are also called class 1 adsorbents¹⁶ and among them small amine molecules, such as tetraethylenepentamine (TEPA),^82,85–87 or amino polymers (e.g. polyethylenimine, PEI)^39,88–90 within robust porous support largely used for CO₂ capture from air. In the chemical grafting method, amine molecules containing the silane group are grafted on the surface of porous solid support via covalent bonding. Grafting occurs through a condensation reaction between the aminosilane species (such as 3-aminopropyltriethoxysilane, APTES) and the hydroxyl group present on the support surface. In general, the synthesis involves mixing the desired amount of porous solid support in a solvent (in general anhydrous toluene) and aminosilane, followed by vigorous stirring and reflux.⁹ The resulting solution is eventually filtered after completion of the reaction and repeatedly washed to remove any unreacted amine molecules from the sorbent.⁸³ Typically, amine-grafted adsorbents are called class 2 adsorbents, which are more resilient than class 1 adsorbents. However, their CO₂ uptake capacity is limited as the surface becomes saturated with aminosilane species.⁹¹ Combining the advantages of impregnation and grafting, a new class of amine-based adsorbents is explored, known as class 3 amine-based adsorbents.¹⁶ Class 3 adsorbents are fabricated by in situ ring-opening polymerization of amine-containing monomers such as aziridine in the pores of support; thus, the resulting aminopolymers are covalently bound to the support, leading to the high stability feature of these materials.⁹² However, the rapid reaction rate of aziridine polymerization makes it challenging to control the polymer formed, and the toxic nature of aziridine makes this approach unappealing on the laboratory scale. Hyperbranched aminosilicas^93–95 and melamine dendrimer-mesoporous silica hybrids,⁹⁶ polysilsesquioxanes^97,98 are a few class 3 adsorbents widely explored in literature for CO₂ capture application. Another type of amine-based adsorbent is called class 4 adsorbent.⁹⁹ They are synthesized by combining both grafting and impregnation techniques, typically known as double-functionalized methods, wherein previously grafted amine-based sorbents are impregnated with aminosilanes to increase their adsorption sites by occulting more amine species to them. However, very limited work has been carried out on class 4 adsorbents for DAC application compared to flue gas CO₂ capture.⁹⁹

3 Evaluation parameters and CO₂ adsorption performance of amine-based solid adsorbent under DAC condition

The selection of suitable adsorbent properties for DAC is cumbersome, as several operational parameters affect the CO₂ capture performance of a sorbent while working in practical environmental conditions. The following sections describe in detail the technical and economic factors affecting the performance of the adsorbent for CO₂ capture from the air.

3.1 CO₂ adsorption performance of amine-based solid adsorbent for DAC

3.1.1 Amine-based solid adsorbent synthesized by physical impregnation method. Low amine stability due to amine degradation and amine evaporation are the key limitations that can obstruct the use of adsorbents having physically impregnated amines in DAC, highlighting the challenges in this task.^88,101–103 However, a few studies on low-volatile high-molecular-weight amines have shown promising progress in addressing this issue. For instance, impregnating polyethylenimine (PEI) molecules with an average molecular weight (M_w) between 400 and 25 000 Da into sorbents maximizes the effective content of amine in the adsorbent.⁸¹ Notably, the literature reported that adsorbents with branched PEI (M_w of 800 Da) achieved the highest CO₂ adsorption capacity (Table 1).^81,88,39 Further, the vacuum level during the drying step is vital to efficiently migrate the amine into the pore network and create favorable amino polymer distributions. Kwon et al.³⁹ reported that a mild vacuum (33.33 mbar) would be better for superior amine loading and amine migration into the pore network. It was seen that the amine efficiency increases with amine content as more free amines are added to the system; however, the amine efficiency will reach a maximum, typically at a loading of 1.6 g PEI per g SIO₂ (corresponds to ∼40 nm aminopolymer film thickness), as the pores eventually become occluded and CO₂ diffusion is hindered. Therefore, a delicate balance between the presence of adsorption sites and amine loading is created to maximize the CO₂ uptake capacity with faster kinetics.¹⁰⁴ To address the thermal stability issue of PEI-modified adsorbents, Choi et al.⁸⁸ studied a 3-aminopropyltrimethoxysilane (APTES) modified PEI silica sorbent (A-PEI) and titanium(IV) propoxide, which enhanced the kinetics, CO₂ uptake capacity (>2 mmol g⁻¹) along with the thermal stability as compared to the pristine PEI modified adsorbents (CO₂ capacity: 1–2 mmol g⁻¹). Modification of the chemical structure of PEI by Ti additives differs in its accessibility or mass transport through the pores. Further, the mesoporous framework is also an important aspect influencing the performance of amine-modified adsorbents for CO₂ capture at very low pressures. Chen et al.¹⁰⁵ reported a PEI-impregnated resin (HP20/PEI-50) with an excellent CO₂ uptake capacity of 2.26 mmol g⁻¹ and almost stable amine efficiency of ∼0.19–0.20 mmol of CO₂ per mmol of N in the PEI content range of 40–50% under DAC conditions. The presence of mesopores (43–68 nm) in the adsorbent facilitated the high amine loading (up to 50 wt%), resulting into enhanced CO₂ diffusion and increased CO₂ adsorption capacity. Similarly, Wang et al.¹⁰⁶ reported a PEI impregnated mesoporous carbon (MC) adsorbent, where the MC framework can accommodate high PEI content owing to the large mesoporous volume and area. The interconnecting channels also serve as a rapid CO₂ route, boosting the sorption/desorption kinetics. In addition, they also enhanced the kinetic inhibition of CO₂ internal diffusion within the PEI films by adding polymer-based surfactant such as sorbitan monooleate (Span 80). This transforms the PEI film from a homogenous solid block to a polymer network, resulting in high CO₂ diffusion rates. Sehaqui et al.¹⁰⁷ reported a highly porous bio-based sorbent composed of oxidised nanofibrillated cellulose (NFC) and 44 wt% polyethylenimine (PEI) synthesized by freeze-drying method to achieve a CO₂ capacity of 2.22 mmol g⁻¹ with stability over five cycles, and exceptionally low adsorption half time of 10.6 min under atmospheric conditions (moist air with ∼400 ppm of CO₂) and at 80% RH. Further, temperature and pressure/concentration also have a greater role in CO₂ adsorption capacity and amine mobility inside the porous adsorbents. Pang et al.¹⁰⁸ explored the mechanism of CO₂ adsorption in amine-based adsorbents by comparing the isotherm at various temperatures and pressures in a dry environment. The group reported a transition of CO₂ sorption isotherm from a transport-limited regime to a thermodynamically limited one. In the case of linear poly(propylenimine) based SBA-15 sorbents, at low partial pressures relevant to direct air capture (P_CO2 = 0.3 Torr), the CO₂ uptake capacity and amine efficiency decreased with the increase in adsorption temperature, in accordance with a thermodynamically limited exothermic adsorption process. However, as the pressure increases (∼1 bar), the CO₂ uptake capacity and amine efficiency of PEI/SBA-15 increased with increasing temperature (cross-over), which may be attributed to the increased flexibility and mobility of the PEI molecules, thus helping greater penetration of CO₂ into the amine sites. Furthermore, when the temperature rises, the films produced by the crosslink between CO₂ and PEI (formation of ammonium carbamate ion pairs) become less dominating over CO₂ penetration, resulting in increased capacity.¹⁰⁹ In the similar context, recently Miao et al.⁸⁶ reported the favourable desorption temperature for PEI and TEPA based SBA-15 adsorbent for DAC application. They found that, a temperature interval of 90–110 °C can be considered as the optimal desorption temperature of PEI/SBA-15 and 90 °C for TEPA/SBA-15 composite (50% amine loading), which could help to minimise amine volatilization. Notably, TEPA/SBA-15 composite exhibited better desorption kinetics than PEI/SBA-15 at the same degassing conditions, evidenced its suitability for DAC.

Table 1 CO₂ capture by amine-functionalized solid adsorbents synthesized by physical impregnation

Support	Amine type	Temp (°C)	CO2 concentration (ppm)	Amine loading (mmol of N per g of the adsorbent)	CO2 adsorption capacity (mmol g^−1)		Amine efficiency (mmol of CO2 per mmol of N)	Method	Ref.
					Dry CO2	Humid CO2
a 67% relative humidity. b 19% relative humidity. c ca. 0.9–1.4 vol% relative humidity. d 50% relative humidity. e 30% relative humidity. f 85% relative humidity. g 80% relative humidity. h 32% relative humidity, TGA, thermogravimetric analysis; IR, infrared analyzer; TPD, temperature-programmed desorption; GC, gas chromatography, and vol., volumetric. Com silica, commercial silica; APTES, (3-aminopropyl)triethoxysilane; PEI, branched polyethylenimine (MW = 800 Da); PEI-A, branched polyethylenimine (MW = 600–1000 Da); PEI-M, branched PEI (MW = 1800 Da); PEI-U, branched polyethylenimine (MW = 2000 Da); PEI-H, branched PEI (MW = 25 000 Da); PEI-ln, linear PEI (Mn = 423 Da); PEI-lnH, linear PEI (MW = 2500 Da); TEPA, tetraethylenepentamine; PAA, poly(allylamine); PEG, poly(ethylene glycol) (Mw ∼ 200); ED, ethylenediamine; PPI, linear poly(propylenimine); PEG poly(ethylene glycol); PGA, poly(glycidyl) amine, (Mn = 50); PEHA, pentaethylenehexamine; PO, propylene oxide; BO, α-butylene oxide; TETA, triethylenetetramine; CTMA^+, cetyltrimethylammonium cations; NOHM-I-PEI, polyethylenimine tethered silica nanoparticles with ionic bond; SIP, solvent impregnated polymers; TTE, triglycidyl trimethylolpropane ether; CA, cellulose acetate.
Com silica	PEI	25	400	10.5	2.36	—	0.22	TGA	88
Com silica	PEI + APTES	25	400	10.7	2.26	—	0.21	TGA	88
Com silica	PEI + titanium(IV) propoxide	25	400	10.5	2.19	—	0.21	TGA	88
Com silica	TEPA	35	400	10.1	2.50		0.25	TGA	87
Fumed silica	PEI-H	25	420	33 wt%	1.18	1.77^a	—	IR	123
Fumed silica	PEI-H	25	420	50 wt%	1.71	1.41^a	—	IR	123
Fumed silica	PEI-ln	25	400	50 wt%	2.34	—	—	IR	81
Fumed silica	PEI	25	400	50 wt%	2.44	—	—	IR	81
Fumed silica	PEI-M	25	400	50 wt%	1.69	—	—	IR	81
Fumed silica	PEI-H	25	400	50 wt%	1.67	—	—	IR	81
Hierarchical silica	PEI	30	400	2.62 g g^−1	2.6	3.36^b	0.15	TGA	39
Mesocellular silica foam	PAA	25	400	7.24	0.86	—	0.12	TGA	89
Mesocellular silica foam	PEI	25	400	10.7	1.74	—	0.16	TGA	89
Mesocellular silica foam	PEI-lnH	25	400	11.4	1.05	—	0.09	TGA	89
Mesoporous silica	PEI-M	20	393.5	40 wt%	—	1.66^c	—	IR	124
γ-Alumina	PEI	25	400	11.2	1.74		0.16	TGA	125
γ-Alumina	PEI	30	400	7.95	—	1.96^d	0.25	IR	126
PME/CTMA^+	PEI	25	400	40 wt%	2.18	2.92	—	TGA	90
SBA-15	PEI	25	400	9.23	1.05	—	0.11	TGA	125
SBA-15	PEI-ln	75	400	50 wt%	0.51	—	—	TPD	127
SBA-15	PEI + PEG	30	400	5.75	0.79	—	0.14	TGA	128
SBA-15	PPI	35	400	13.8	1.25	—	0.16	TGA and vol.	108
SBA-15	PEI	25	400	75 wt%	1.90	—	—	TGA	86
SBA-15	TEPA	25	400	75 wt%	3.44	—	—	TGA	86
SBA-15	PPI	35	400	8.17	0.31	—	0.04	TGA	129
SBA-15	PGA	35	400	10.2	0.6	—	0.11	TGA and vol.	109
SBA-15	PEI	30	400	30 wt%	0.65	—	0.10	TGA	130
SBA-15	PPI (linear)	30	400	14	1.85	—	0.13	TGA	131
SBA-15	TETA	35	400	19	3.50	—	0.18	TGA	132
Zr-SBA-15	PEI	25	400	8.3	0.85		0.10	TGA	133, 134
Ph-3-SBA-15	ED	35	400	16.7	1.9	2.9^e	0.11	TGA	41
CA-SiO2	PEI	35	395	16.1	0.59	1.60^f	—	TGA	101
Sipernat 50S	PEHA-PO-1-2	25	400	—	1.25	—	—	TGA	40
Sipernat 50S	PEHA-BO-1-2	25	400	—	1.11	—	—	TGA	40
Sipernat 50S	TEPA-PO-1-2	25	400	—	1.34	—	—	TGA	40
Resin HP20	PEI	25	400	50 wt%	2.26	—	—	—	105
Mesoporous carbon	PEI	25	400	55 wt%	2.25	2.58^g	—	GC	106
MIL-101(Cr)	PEI	25	400	1.76	1.25	1.26^h	0.10	TGA	135
MIL-101(Cr)	PEI	25	400	50 wt%	1.81	—	0.14	vol.	43
MIL-101(Cr)	TEPA	25	400	50 wt%	2.14	—	0.16	vol.	43
TTE	PEI-H	20	400	30 wt%	—	1.14	—	Isotopic analyzer	136
Nanofibrillated cellulose (NFC)	PEI-A	25	400	44 wt%	—	2.22	0.22	IR	107
NOHM-I-PEI in SIP film	PEI-U	25	400	50 wt%	1.05	1.66	—	TGA & IR	110

---

Several next-generation encapsulation techniques, such as micro-encapsulated carbon sorbents (MECS) and solvent-impregnated polymers (SIPs), have been recently explored to develop hybrid CO₂ capture sorbents that combine the benefits of solid sorbents and liquid solvents. The kinetics of CO₂ adsorption in a liquid solvent is usually slow. As a result, the liquid solvent is kept inside the polymer matrix, which has a high CO₂ permeability, allowing for a synergistic effect on CO₂ sorption behaviors, increased mass transfer, and rapid CO₂ capture. The advantage of SIPs over MECS is that it does not require a microfluidic device to create sorbent particles, making them more straightforward and having a better potential for scale-up. Recently, Rim et al.¹¹⁰ synthesized liquid-like polyethylenimine tethered silica nanoparticles with an ionic bond (NOHM-I-PEI), incorporated into a shell material and UV-cured to yield gas-permeable solid sorbents with uniform NOHMs loading (NPEI-SIPs) as shown in Fig. 6. NPEI-SIPs had a CO₂ adsorption capacity of 1.05 mmol g⁻¹ at 400 ppm, 25 °C at 50% PEI loading, with rapid CO₂ adsorption kinetics (4.16 μmol g⁻¹ min⁻¹) and claimed to be a well-suited system for DAC application.

Fig. 6 Structure of liquid-like nanoparticle organic hybrid materials with ionically tethered polyethylenimine (NOHM-I-PEI) and the schematic of microdroplets of NOHM-I-PEI in solvent impregnated polymers (SIPs). Reproduced with permission.¹¹⁰ Copyright 2021, Wiley-VCH GmbH.

3.1.2 Amine-based solid adsorbent synthesized by chemical grafting method. In general, amine-based adsorbents synthesized by the chemical grafting method displayed high stability and easy regeneration during CO₂ adsorption–desorption cycles, even in humid conditions, compared to those obtained by the physical impregnation method.^111–114 Due to the covalent bonding between the amines and silica supports, the chemical grafting process can produce amine–silica composites that are more thermally stable than those formed via a physical impregnation method.⁸³

Recently, amine-appended MOFs have been explored to show remarkable performance in DAC owing to their extraordinary surface area, giant pore volume, and tunable pore architecture, allowing amines to immobilize into the pore wall and interact with the CO₂ at ultra-dilute concentration. For instance, Lee et al.¹¹⁵ prepared an amine-appended metal–organic framework (en-Mg₂(dobpdc)) (where H₄-dobpdc = 4,4′-dihydroxy-(1,1′-biphenyl)-3,3′-dicarboxylic acid) functionalized with en (ethylene diamine). This material exhibited a superior CO₂ adsorption capacity of 2.83 mmol g⁻¹ from ambient air (p_CO2 = 0.4 mbar) owing to carbamic acid (not a carbamate) formation by the chemical reaction associated with the free amine group and CO₂. However, the CO₂ sorbent shown by Lee et al. could not achieve saturation at 400 ppm CO₂ in air, which was attributed to strong intramolecular interactions between the neighboring amine groups, which can be surpassed at higher pressure CO₂.^116,117 As a consequence, the CO₂ uptake of en-Mg₂(dobpdc) drops sharply to 0.120 mmol g⁻¹ at a somewhat higher temperature of 50 °C at 0.4 mbar. To overcome such a problem, Liao et al.¹¹⁸ substituted the ‘ethylenediamine (en)’ molecule with an even shorter diamine named ‘hydrazine (N₂H₄)’; where with an overall amine loading of 6.01 mmol of N per g of the adsorbent, a very high CO₂ adsorption capacity of 3.89 mmol g⁻¹ amongst other MOFs was achieved (Table 2). In addition, Gebald et al.¹¹⁹ reported a CO₂ adsorption capacity of 1.11 (dry CO₂), and 2.13 (humid CO₂, 91% relative humidity) over APDES–NFC (APDES: 3-aminopropylmethyldiethoxysilane, NFC: nano fibrillated cellulose) at 0.4 mbar. The remarkable CO₂ adsorption performance of APDES–NFC was presumably attributed firstly to the interaction of mesopores of APDES–NFC with CO₂via physisorption and hence increased amine efficiency (0.54 mmol of CO₂ per mmol of N) is achieved. Secondly, the aminosilane dimer molecule for each hydroxyl group at the cellulose surface and unreacted silanol groups presumably catalyzed the interaction of CO₂ molecules with the amine groups, enhancing the amine efficiency beyond the theoretical limit of carbamate formation. Lu et al.¹²⁰ examined the impact of amine-grafted PPN-6-CH₂DETA (PPN: porous polymer networks; DETA: diethylenetriamine) on CO₂ capture from the air. High CO₂ selectivity and uptake capacity (1.04 mmol g⁻¹) was observed due to the highly porous three-dimensional carbon scaffolds. Notably, tertiary amine DETA (diethylenetriamine) has a 6% higher CO₂ loading capacity with low heat of adsorption (53.8 vs. 67.3 kJ mol⁻¹) as compared to the amine-modified pore expanded silica (TRI-PE-MCM-41) at 400 ppm (0.98 mol kg⁻¹).

Table 2 CO₂ capture by amine-functionalized solid adsorbents synthesized by chemical grafting and in situ polymerization process

Support	Amine type	Temp. (°C)	CO2 concentration (ppm)	Amine loading (mmol of N per g of the adsorbent)	CO2 adsorption capacity (mmol g^−1)		Amine efficiency (mmol of CO2 per mmol of N)	Method	Ref.
					Dry CO2	Humid CO2
Mesocellular silica foam	APTMS	25	400	2.70	0.54	—	0.20	TGA	84
Mesocellular silica foam	APTMS	25	400	3.75	1.0	—	0.30	TGA	84
Mesocellular silica foam	MAPS	25	400	2.41	0.17	—	0.07	TGA	84
O-Al2O3	APTES	30	400	5.43	0.76	—	0.14	TGA	91
D-Al2O3	APTES	30	400	4.13	0.62	—	0.15	TGA	91
Silica gel	AEATPMS	25	400–440	2.48	0.40	0.44^a	Dry: 0.16	TGA	112
							Humid: 0.18
Nano fibrillated cellulose	AEAPDMS	25	506	4.9	—	1.39^a	0.28	IR	113
Nano fibrillated cellulose	APDES	23	400	—	1.11	2.13^b	Dry: 0.26	IR	119
							Humid: 0.51
PPN-6	DETA	22	400	8.5	1.04		0.12	vol.	120
RFAS	APTES	25	400	8.07		1.71^c	0.21	IR	137
SBA-15	APTMS	25	395	—	0.14		—	TGA	138
SBA-15 pellet	APTMS	25	395	—	0.09	0.13^d	—	TGA	138
SBA-15	Alkyl halide + ammonia	30	400	1.62	0.07	—	0.043	TGA	75
SBA-15	TEPA	23.4	400	12.2		3.59^e	—	TPD	82
Pore-expanded MCM-41	TRI	30	400	—	0.64	0.55^f	—	TGA	103
Pore-expanded MCM-41	TRI	25	400	7.9	0.98	—	0.12	TGA	33
Mg2(dobpdc)	Ethylenediamine	25	390	—	2.83	—	—	vol.	115
Mg2(dobdc)	Hydrazine	25	400	6.01	3.89	—	—	TGA	118
Mg-MOF-74	Ethylenediamine	25	400		1.51	—	—	TGA	139
Mg2(dobpdc)	mmen	25	390	—	2.0	—	—	vol.	42
Cr-MIL-101-SO3H	TAEA	20	400		1.12	—	—	TGA	140

---

In situ polymerization
a 40% humidity. b 91% relative humidity. c 4% relative humidity. d 80% relative humidity. e 49% relative humidity. f 73% relative humidity. g 60% relative humidity. h Fully humidified condition; TGA, thermogravimetric analysis; IR, infrared analyzer; TPD, temperature-programmed desorption; GC, gas chromatography, and vol., volumetric. APTES, (3-aminopropyl)triethoxysilane; MAPS, N-methylaminopropyl-trimethoxysilane; AEAPDMS, [N-(2-aminoethyl)-3-aminopropylmethyl]dimethoxysilane; APDES, 3-aminopropylmethyldiethoxysilane; APTMS, (3-aminopropyl)trimethoxysilane; AEATPMS, [N-(2-aminoethyl)-3-aminopropyl]trimethoxysilane; DETA, diethylenetriamine; mmen, N,N′-dimethylethylenediamine; TAEA, tris(2-aminoethyl)amine; TRI, 2-[2-(3-trimethoxysilylpropylamino)ethylamino]ethylamine, AHTSA, amine hybrid titania/silsesquioxane composite aerogel.
SBA-15	Z-L-Lysine + APTMS	25	400	5.18	0.60	—	0.12	TGA	92
Hybrid silica	APTES	30	400	4.5	—	1.68^g	0.37	TGA	111
SBA-15	Aziridine	25	400	9.9	—	1.72^h	0.17	MS	95
AHTSA	APTES	30	400	8.47	1.64	—	0.19	IR	114
Macroporous silica	L-Alanine	50	400	10.98	2.65	—	0.24	TGA	141

---

3.1.3 Amine-based solid adsorbents synthesized by in situ polymerization method. Jones et al.⁹⁵ reported the first study on a new type of amine-based adsorbent (hyperbranched aminosilica (HAS)), synthesized by in situ polymerizations of amine-containing monomers (aziridine) in the porous SBA-15 support. The HAS contains aminopolymers covalently tethered to the support, leading to high amine loading and regeneration stability as compared to the grafted adsorbents, in which the aminosilane is linked to other already grafted aminosilanes by Si–O–Si bond and is thus indirectly linked to the support.⁸³ Choi et al.⁹⁵ also reported the facile one-pot synthesis of aziridine-based amine-based silica SBA-15. The resultant material with extraordinary amine loading of 9.9 mmol of N per g exhibited 1.72 mmol g⁻¹ CO₂ uptakes at 400 ppm under a fully humidified environment. They observed that functionalization (amine inclusion) happened in the pores at lower amine loadings until a critical diameter was reached. Further polymerization occurred most likely at the pore mouth, which resulted in pore blockage with no change in the pore diameter. In the same context, Qi et al.¹²¹ synthesized a hybrid adsorbent by grafting alkyliodine as the initiator, followed by growing the amino polymer on the mesoporous support via the cationic polymerization of 2-methyl-2-oxazoline and subsequent acid hydrolysis to produce a high density of grafted linear PEI-like polymer composed of secondary amines with terminal primary amines. Even with high amine loadings, the hybrid adsorbent demonstrated excellent amine efficiency (>0.8 mmol of CO₂ per mmol of N), resulting in an exceptional CO₂ absorption capacity (12 mmol g⁻¹) under simulated flue gas conditions (8% CO₂ with a relative humidity of 18% at 25 °C). Clearly, amine-based adsorbents synthesized through in situ polymerization offer a better advantage over amine-impregnated or amine-grafted adsorbents, and further exploration for DAC application is necessary, as their bulk production is easy and energy-efficient.¹¹¹ Recently, Lyu et al.¹²² reported the first amine functionalized covalent organic frameworks (COFs) termed COF-609 by incorporation of an aliphatic amine such as tris(3-aminopropyl)amine into the framework. The CO₂ uptake capacity of COF-609 was found to be 0.393 mmol g⁻¹ under simulated air containing 400 ppm CO₂ (50% RH, 25 °C and 1 atm), which is more than dry CO₂ uptake, demonstrating the efficiency of COF-609 for efficient CO₂ uptake from very diluted sources under humid condition. The DAC performance of various amine-based adsorbents synthesized by in situ polymerization techniques is listed in Table 2, and the CO₂ uptake capacity and amine efficiency of some selected amine-based adsorbents are shown in Fig. 7.

Fig. 7 Comparative CO₂ adsorption performance of selected amine-based adsorbents (based on Tables 1 and 2).

3.2 CO₂ adsorption kinetics of amine-based solid adsorbent for DAC

Adsorption kinetics is a time-dependent calculation of adsorption at constant pressure or concentration, which provides the necessary information about the adsorbate diffusion into the adsorbent's pores. Since CO₂ concentration in air is very low (400 ppm), the CO₂ adsorption kinetics (capture-release cycle) of the amine-based adsorbents is worth exploring. The adsorption kinetics also affects the working capacity of the adsorbent, and the working capacity can be the same as the equilibrium adsorption capacity of the sorbent possesses high adsorption kinetics.⁸⁸ Despite the significant importance of adsorption kinetics, the number of publications reporting the kinetic data of CO₂ during direct air capture is currently limited. Only a few articles describe the sorbent's kinetic behavior in terms of adsorption half time^{95,111,113,114,134,142} or the breakthrough time.¹⁴³ Adsorption half time is the time to reach half of the pseudo-equilibrium adsorption capacity, typically calculated from pure component isotherm either by volumetric or gravimetric measurement.⁹⁵ Compared to the flue gas environment (10–15% CO₂), the adsorption kinetics of CO₂ in DAC (0.04% CO₂) typically take a longer time for amine-based adsorbents.³⁹ This behavior is primarily affected by various factors such as CO₂ diffusion to bulk materials (boundary layer diffusion), mass transfer in and out of the pores, structure and density of the grafted amines, accessibility to amine moieties, intrinsic chemical reaction rates, and others.^39,94

For instance, Kwon et al.³⁹ compared the kinetics of conventional mesoporous material-supported amine sorbents (class 1, PEI-based sorbents) during CO₂ adsorption from the simulated air and flue gases (400 ppm and 10% CO₂) at a fixed temperature. At higher CO₂ concentration, 2.6 g PEI per g sorbent exhibited 30–40% higher CO₂ uptake at the initial stage but it lasted only for a few minutes, whereas the same material showed 70–80% of the total CO₂ uptake capacities at a slow initial rate for 400 ppm CO₂ concentration. The observed trend can be attributed to a higher driving force for adsorption under flue gas conditions causing an enhanced surface diffusion barrier during the initial stage of the adsorption process. In contrast, for 400 ppm CO₂, a low driving force favoured diffusion of CO₂ relatively deeper into the aminopolymer film; hence, adsorption occurred until saturation was achieved. For instance, PEI-H-SiO₂ displayed a high uptake capacity of 2.6 mmol CO₂ per g at 400 ppm dry CO₂ at 50 °C.³⁹ Though the temperature-induced adsorption capacity (kinetic limitation) was observed to be improved at 400 ppm CO₂, it was not accompanied by increased adsorption kinetics as the sorption rates decreased with the increase in temperature, conversely to the case of the 10% CO₂ conditions.

The CO₂ adsorption kinetics greatly depends upon the synthesis methodology of amine loaded into the adsorbent. It has been seen that compared to amine-impregnated adsorbents, amine-grafted and in situ polymerized amine-based adsorbents exhibit better CO₂ adsorption kinetics in DAC conditions except for NFC–PEI (made by a freeze-drying process) (Fig. 8, Table 3). For instance, Choi et al.⁹⁵ reported the adsorption half time of various amine-modified hyperbranched aminosilica (HAS) adsorbents and found the CO₂ adsorption capacity of amine-modified HAS adsorbents was increased to 1.72 mmol g⁻¹ by increasing the amine loading to 9.9 mmol g⁻¹ with an adsorption half-time of ∼167 min. Later, the same group also compared the adsorption kinetics of various amine-based adsorbents (class 1 and class 3) for CO₂ capture from air.⁸⁸ They found that the conventional class 1 amine-based adsorbent (PEI/silica) also displayed the highest CO₂ uptake of 2.36 mmol g⁻¹ when exposed to CO₂ (400 ppm) in Ar, due to the presence of a larger amount of polymeric amines inside the pores (10.5 mmol N per g of adsorbent). However, it led to a longer adsorption half-time, whereas, after mixing the additive with PEI (tetrapropylorthotitanate and (3-aminopropyl)triethoxysilane (APTES)), the adsorption half time of the resultant mixture was reduced without significant change in CO₂ uptake. The adsorption half time and corresponding average CO₂ adsorption rate was 309 min and 3.8 μmol g⁻¹ min⁻¹ for PEI/silica, 196 min and 5.8 μmol g⁻¹ min⁻¹ for A-PEI/silica, 210 min and 5.2 μmol g⁻¹ min⁻¹ for T-PEI/silica and 167 min and 5.2 μmol g⁻¹ min⁻¹ for HAS6 respectively.⁸⁸ The reduced CO₂ adsorption rate was probably due to a change in the chemical structure of PEI by mixing with additives, causing tuned amine accessibility and mass transport through the pores. A study of air capture using TRI-PE-MCM-41 (triamine-grafted pore-expanded mesoporous silica) recorded a breakthrough time of 167 min and an adsorption capacity of 0.98 mmol g⁻¹.¹⁴³ In another study using MAHSM (mono amine-based hybrid silica material), the CO₂ adsorption capacity of 1.68 mmol g⁻¹ was obtained with the adsorption half time of 50 min. A rate of 17.9 μmol g⁻¹ min⁻¹ during CO₂ capture from the simulated air (containing 400 ppm CO₂, 60% RH) was achieved at 30 °C with a flow rate of 100 ml min⁻¹ in 24 h.¹¹¹ Gebald et al. reported the adsorption half time of AEAPDMS-NFC-FD (aminosilane modified nanofibrillated cellulose aerogel) was 92 min, corresponding to an average CO₂ adsorption rate of 7.6 μmol g⁻¹ min⁻¹, which is quite less than PEI.¹¹³ On the other hand, Sehaqui et al.¹⁰⁷ reported a porous sheet-like biocomposite (PEI modified nanofibrillated cellulose) adsorbent, synthesized by freeze-drying process, exhibiting exceptionally low adsorption half time of 10.6 min with a faster uptake rate (105 μmol g⁻¹ min⁻¹) due to the higher porosity and micro meter range pores available in these sorbents.

Fig. 8 Comparative CO₂ adsorption kinetics of amine-based adsorbents (based on Table 3).

Table 3 CO₂ adsorption kinetics of amine-functionalized solid adsorbents for DAC

Amine based composite	CO2 adsorption capacity (mmol g^−1)	Adsorption halftime (min)	Adsorption rate (μmol g^−1 min^−1)	Method	Ref.
a Data retrieved from data from graph software (version 1.0); Com silica, commercial silica.
Com silica + PEI	2.36	309	3.8	TGA	88
Com silica + PEI + APTES	2.26	196	5.8	TGA	88
Com silica + PEI + titanium(IV) propoxide	2.19	210	5.2	TGA	88
SBA-15 + aziridine	1.72	167	—	MS	95
Hybrid silica + APTES	1.68	50	17.9	TGA	111
Zr-SBA-15 + PEI^a	0.85	185	—	TGA	134
Ce-SBA-15 + PEI^a	0.68	264	—	TGA	134
Ti-SBA-15 + PEI^a	0.64	210	—	TGA	134
Al-SBA-15 + PEI^a	0.29	171	—	TGA	134
SBA-15 + PEI^a	0.19	288	—	TGA	134
NFC + PEI	2.22	10.6	105	IR	107
NFC-FD + AEAPDMS	1.39	92	7.6	IR	113
AHTSA + APTES	1.64	15.88	—	IR	114
RFAS + APTES	1.78	233	—	IR	137

---

Though amine-impregnated adsorbents are not the best in terms of CO₂ adsorption kinetics, exceptions are the amine-impregnated MOFs, where the CO₂ adsorption kinetics readily depend on the textural properties of MOFs, the presence of the open metal site and associated amine ligands. For instance, in the case of mmen-Mg₂(dobpdc) (where mmem is N,N′-dimethylethylenediamine and dobpdc is 4,4′-dihydroxy-(1,1′-biphenyl)-3,3′-dicarboxylate),⁴² CO₂ adsorption kinetics of mmen-Mg₂(dobpdc) tend to be slightly faster than other amine-based adsorbents, where amines are deposited on the adsorbents by evaporation or polymerization method. Notably, mmen-Mg₂(dobpdc) took only 60 min to achieve the 4.6% equilibrium CO₂ adsorption capacity compared to the benchmark impregnated polyethylenimine silica gel under DAC condition,⁸⁸ which took nearly 210 min to reach the same equilibrium adsorption capacity. The readily accessible amines in mmen considerably improve adsorption rates, hence facilitating the rapid adsorption–desorption cycles. A similar relationship of CO₂ adsorption kinetics with the textural properties and molecular weight of amine can be seen in silica-based adsorbents. Although both the composites (PPI/MCF and PPI/SBA-15) (where, PPI is poly(propylene imine), MCF is a mesocellular silica foam and SBA-15 is Santa Barbara Amorphous-15, a type of mesoporous silica) adsorb roughly the same amount of CO₂, PPI/MCF, having a larger pore volume, has resulted in more rapid uptake under the adsorption conditions of 400 ppm CO₂/N₂ at 35 °C.¹⁰⁸ A rapid tailing out for secondary amines is reported, while primary amines displayed CO₂ adsorption for a prolonged time at a faster initial rate. To study the effect of the molecular weight of adsorbents and their CO₂ adsorption, PPI polymers of different molecular weights were prepared by altering the polymerization time. The lower composite molecular weight oligomer showed faster CO₂ adsorption kinetics than the higher molecular weight polymer composite and vice versa. This is due to the relative degree of amine branching that affects the CO₂ affinity and not just the diffusion properties of the gas.¹²⁹ After all, the CO₂ adsorption kinetics of amine-based adsorbents also depend upon the design of the adsorbent bed (fluidized bed or small-scale fixed bed). For instance, Zhang et al.¹²⁴ used a bubbling fluidized bed reactor for direct air CO₂ capture on PEI–silica adsorbent and found that almost 100% equilibrium CO₂ adsorption capacity can be accomplished with 7.5 s contact time between the adsorbent and the ambient air. This might be due to the adsorbent's rapid reaction kinetics, more turbulent gas flow to the adsorbent in the reactor, and highly efficient gas–solid contact in the whole riser of the reactor.

3.3 Stability of amine-based solid adsorbents for DAC

The stability of CO₂ adsorbents is critical for practical DAC application, as the lifetime of adsorbents has a significant influence on the entire process cost.⁹² The degree of change in adsorption capacities across repeated adsorption/desorption cycles is described as the regenerability of CO₂ adsorbents.⁸⁸ A fully regenerable sorbent can resist several adsorption/desorption cycles without losing its CO₂ adsorption capacity, reducing the frequency of sorbent replacement and hence improving the DAC economics. From the practical point of view, amine-based adsorbents are typically subjected to thermal, moisture, and oxidative stability issues, which are briefly discussed in this section. The sorbent regeneration and cyclic CO₂ adsorption–desorption stability of amine-based solid adsorbents are described in Table 4.

Table 4 Sorbent regeneration and cyclic CO₂ adsorption–desorption stability of amine-based solid adsorbents

Support	Amine type	Sorbent regeneration condition	Stability performance	Ref.
a Data retrieved from data from graph software (version 1.0).
Silica	PEI	110 °C for 3 h under Ar flow at 100 mL min^−1	∼30% capacity loss in 4 cyclic runs	88
Silica	A-PEI	110 °C for 3 h under Ar flow at 100 mL min^−1	∼9% capacity loss in 4 cyclic runs	88
Silica	T-PEI	110 °C for 3 h under Ar flow at 100 mL min^−1	∼1.3% capacity loss in 4 cyclic runs	88
Fumed silica	PEI	85 °C for 3 h under vacuum	∼3.5% capacity loss in 4 cyclic runs	123
Fumed silica	PEI	85 °C for 3 h under airflow at 335 mL min^−1	No apparent loss of capacity in 4 cyclic runs	81
Hierarchical silica	PEI	110 °C for 6 h under He flow at 100 mL min^−1	16% capacity loss in 5 cyclic runs in humid conditions, whereas no appreciable change in CO2 adsorption capacity in dry condition	39
Mesoporous silica	PEI	130 °C for 2 h under N2 flow (moisture content 8.8 vol%) at 8000 mL min^−1	No apparent loss of capacity in 4 cyclic runs	124
SBA-15	PEI	110 °C for 9 h under Ar flow at 100 mL min^−1	Stable over short multicycle operations (3 cyclic runs)	125
SBA-15	PPI	110 °C for 10 min under N2 flow	No apparent loss of capacity in 50 cyclic runs	108
SBA-15	LPPI	110 °C for 30 min under He flow at flow rate of 90 mL min^−1	No apparent loss of capacity in 20 cyclic runs	131
Zr-SBA-15	PEI	110 °C for 3 h under Ar flow at 100 mL min^−1	2% capacity loss in 4 cyclic runs	134
Ph-3-ED/SBA-15	Ethylenediamine	90 °C for 10 min under He flow at flow rate of 90 mL min^−1	No apparent loss of capacity in 25 cyclic runs	41
CA-SiO2	PEI	110 °C for 2 h under He flow	∼9% pseudo-equilibrium capacity and 16% breakthrough capacity loss in 20 cyclic runs	101
Resin-HP-20	PEI	100 °C for 3 h under N2 flow at 100 mL min^−1	∼1.5% capacity loss in 5 cyclic runs	105
Mesoporous carbon	PEI	110 °C under N2 flow at 50 mL min^−1	3% capacity loss in 10 cyclic runs	106
Mg/(dobdc)	Ethylenediamine	110 °C for 3 h under Ar flow	No apparent loss of capacity in 4 cyclic runs	139
Mg2(dobpdc)	Ethylenediamine	150 °C for 2 h under simulated air (0.39 mbar CO2) purge at 60 mL min^−1	6% capacity loss in 5 cyclic runs	115
Mg2(dobpdc)	mmen	150 °C for 30 min under N2 flow at 25 mL min^−1	No apparent loss of capacity in 10 cyclic runs	42
MIL-101-Cr	PEI	110 °C for 3 h under He flow at 90 mL min^−1	CO2 uptake dropped 2.7% and 1.9% after the first and second cycles, respectively (total 3 cyclic runs)	135
Cr-MIL-101-SO3H	TAEA	80 °C for 60 min at high vacuum	No apparent loss of capacity in 15 cyclic runs	140
TTE	PEI	120 °C for 30 min by water steam at 2000 mL min^−1	No apparent loss of capacity in 10 cyclic runs	136
Nano fibrillated cellulose	AEAPDMS	90 °C and 40% relative humidity under Ar flow at 800 mL min^−1	Stable over 20 cyclic runs	113
Silica gel	AEATPMS	90 °C under vacuum with a sufficient humid environment	Good repeatability over 40 consecutive cyclic runs	112
Pore-expanded MCM-41	TRI	100 °C for 3 h by synthetic air at 10 mL min^−1	∼24% capacity loss in 4 cyclic runs^a	103
SBA-15	Aziridine	110 °C for 3 h under Ar flow at 100 mL min^−1	Relatively stable for 4 cyclic runs	95
Hybrid silica	APTES	80 °C for 15 min under He flow at 70 mL min^−1	Stable over 50 cyclic runs	111
AHTSA	APTES	90 °C for 30 min under N2 flow at 300 mL min^−1	No apparent loss of capacity in 15 cyclic runs	114
SBA-15	Z-L-Lysine, APTMS	110 °C for 3 h under a pure argon purge	Stable over short multicycle operations (3 cyclic runs)	92

---

3.3.1 Thermal stability of amine-based solid adsorbents for DAC. Amine-based adsorbents are generally heated above 80 °C to desorb the captured CO₂ in a vacuum or under the flow of carrier gas such as air or inert gas. The thermal stability of amine-based adsorbents depends on the volatility (boiling point) of the incorporated amines and their interaction with the support. For instance, amine-impregnated adsorbents are considered to accommodate a large number of amine species and thus have high CO₂ adsorption capacities.^32,83 Still, they are also amongst the least stable adsorbents due to the lack of strong interactions between the supports and the amine species. On the other hand, amine-grafted and in situ polymerized amine-based adsorbents often possess superior thermal stability over hundreds of cycles due to the strong covalent bonds between the amine and supports. For instance, Chaikittisilp et al.⁹² reported that APTMS in situ modified poly(L-lysine) brush–mesoporous hybrid silica exhibited high thermo-cyclic stability due to the strong covalent bonds between poly(L-lysine) and silica support, and the hybrid adsorbent can be regenerated under argon purging at 110 °C. Amine evaporation from the adsorbent and urea formation (dehydrative condensation between amines and CO₂) is a side reaction that could limit the cyclic stability of amine sorbents if the desorption occurred at a higher temperature.⁷⁰ Therefore, in the case of PEI-impregnated adsorbent, PEI with molecular weights higher than 600–800 g mol⁻¹ are more suitable for stable DAC applications due to their low volatility (high boiling point ∼250 °C). Especially, branched PEI with high molecular weight (M_w = 25 000 g mol⁻¹) exhibited no sign of leaching.⁸¹ However, aminopolymer containing primary amines is more susceptible to urea formation than secondary or tertiary amines. Such aminopolymers may get deactivated rapidly in the presence of CO₂ atmosphere through the intermediacy of isocyanate followed by a reaction with primary or secondary amine into di- and tri-substituted open-chain urea or N-substituted imidazolidinones.⁷⁰

The literature reports inferred that incorporating additives to polyamines during adsorbent synthesis and desorption of amine-based adsorbent under humid conditions (as also mentioned in section 3.3.3) is a promising way to prevent urea formation.⁷² Choi et al. reported a comparative study on cyclic CO₂ stability of low molecular weight PEI-impregnated silica before and after incorporating additives titanium(IV) propoxide (tetrapropyl orthotitanate) and (3-aminopropyl)triethoxysilane.⁸⁸ It was observed that after tetrapropyl orthotitanate (C₁₂H₂₈O₄Ti) modification, the amine-based adsorbent exhibited only 1.36% loss in CO₂ cyclic stability compared to PEI-based silica (30% loss). Goeppert et al. also reported the prolonged 50 thermo-cyclic stability of polyamines (PEHA and TEPA) supported Sipernat 50S silica after adding an epoxide stabilizing agent (propylene oxide).⁴⁰ The group reported that compared to TEPA-PO-1-2/50S, PEHA-PO-1-2/50S based adsorbents are more thermally stable, even up to 50 cycles due to the presence of more one –CH₂–CH₂–NH– group than TEPA. A desorption time of 5 min at 85 °C was sufficient to desorb all the CO₂ from the adsorbent with no noticeable changes over time. Abhilash et al.¹¹¹ synthesized monoamine-based hybrid silica material using a combination of 3-aminopropyl triethoxysilane (APTES) and vinyl triethoxysilane (VS) and carried out a CO₂ adsorption experiment under a 400 ppm CO₂ environment at 60% RH.¹¹¹ Fifty cycles were performed at 30 °C in ambient air followed by desorption at 80 °C under He flow, and the shifting of carbamate carbonyl peak (164.5 ppm) was verified using ¹³C CP-MAS NMR. The adsorbent exhibited excellent cyclic stability with no shifting of carbamate carbonyl peak, indicating no urea formation. Interestingly, the studied adsorbent is highly stable and can regenerate quickly by desorbing CO₂ at a relatively low temperature (80 °C) within a few minutes (10 min).

3.3.2 Oxidative stability of amine-based solid adsorbent for DAC. Oxidative degradation is a severe concern for practical direct air capture applications, where the air is used for desorption medium at low to moderate temperatures. Since amine-based adsorbents are exposed to a large concentration of O₂ (21%) present in the air, amine oxidation is likely to occur via a radical-mediated mechanism,¹⁰⁸ eventually degrading the sorbent.^{131,144–147} Oxidative stability also depends upon the type of amines employed in the porous adsorbents.¹⁴⁸ As suggested in the NMR studies conducted by Ahmadalinezhad and Sayari during PEI oxidation, aminopolymers containing only primary amines are the most resistant to oxidative degradation, followed by those containing only secondary amines, and the least stable the highly branched aminopolymers containing all the primary, secondary and tertiary amines.¹⁴⁸ Based on the stability of amine-based adsorbents in the presence of O₂, they could be ranked as follows: primary monoamine (e.g. propylamine) ≈ polyallylamine (PAA) > secondary monoamine (e.g. N-methylpropylamine) > secondary polyethylenimine (e.g., LPEI) > primary + secondary ethyleneamines (e.g. (propyl) diethylenetriamine) > primary + secondary + tertiary polyethylenimine (e.g., BPEI). In another study carried out by Sayari et al.,¹⁴⁹ multiple amines such as primary (pMono: 3-(aminopropyl)trimethoxysilane), secondary (sMono: trimethoxy[3-(methylamino)propyl]silane) and tertiary (tMono: (N,N-dimethylaminopropyl)trimethoxysilane) monoamines along with triamine (TRI: N¹-(3-trimethoxysilylpropyl)diethylenetriamine), containing both secondary and primary amines, were grafted on PE-MCM-41 to investigate their oxidative resistance when exposed to air at higher temperatures (120 °C). They concluded that primary monoamine is more stable than the secondary and the mixed amine-containing materials towards oxygen at high temperatures. On the other hand, polymeric secondary amines and mixed amines undergo irreversible activity loss attributed to the diminished amine basicity due to the formation of imides, amides, and imines.³² In a similar context, Zhang et al. had shown that the CO₂ adsorption capacity of linear polyethylenimine (LPEI5k) impregnated on fumed silica, containing exclusively secondary amines was reduced by 92.8% due to the imine and amide formation after heating at 100 °C for 20 h in the air;¹⁵⁰ whereas, in 2011, Bali et al.¹⁵¹ demonstrated that poly(allylamine) (PAA), which contains only primary amines, is more resistant to oxidative degradation than the PEI-based sorbents.

Various researchers have attempted to enhance the oxidative stability of amine-based sorbents by functionalizing or blending the aminopolymers with certain additives. Chuang et al. used poly(ethylene glycol) (PEG) as an additive to reduce the oxidation of supported TEPA by facilitating hydrogen bonding between PEG and TEPA.^152,153 By restricting amine sites from access to oxygen, these hydrogen-bonding interactions between amines and OH groups boosted TEPA dispersion and reduced amine oxidation to imide species.^152,153 Choi et al.¹⁵⁴ hybridized branched PEI with 1,2-epoxybutane, enabling the availability of hydrogen bonding groups directly onto the aminopolymer to suppress urea formation and oxidative degradation of amines simultaneously by reducing the fraction of primary amine. The oxidative stability of another aminopolymer, namely poly(glycidyl amine) (PGA) containing both ethylene oxide and primary amine groups in each repeating unit, was studied by Sujan et al.¹⁰⁹ The sample was exposed to 21% O₂ with a balance of N₂ at 110 °C, resulted in the enhancement in the stability of the sorbent, and hence the sorbent could retain 50% of its CO₂ adsorption capacity.¹⁰⁹

In a study by Goeppert et al. under oxidative conditions, the stability of tetraethylenepentamine (TEPA) and pentaethylenehexamine (PEHA) modified sorbents were found to be significantly enhanced by introducing propylene oxide (PO) as intramolecular alcohol functionalities.⁴⁰ The oxidative stability experiment conducted in the aerobic atmosphere at 100 °C for 24 h revealed that the CO₂ adsorption capacity of PO containing TEPA sorbent remains stable at 25 °C but decreased by 16% at 55 °C and 18% at 85 °C. For PEHA-PO, comparable results were obtained, indicating the increased oxidative stability of PO-modified aminopolymer sorbents.

Another approach to enhance oxidative stability included using small molecule poly(propylenimine) (PPI) in linear and branched aminopolymer forms. The linear PPI, namely tripropylenetetramine (TPTA) (188 g mol⁻¹) containing primary as well as secondary amines, showed only a 20% loss in amine efficiency after oxidative treatment in air at 110 °C for 24 h.¹³² Pang et al.^108,132 critically analyzed the oxidative stability of small and large molecules of amine introduced to the porous support. Switching from a PEI to a poly(propylenimine) (PPI) (without secondary amine) increases the spacing of the amines in the backbones. It significantly improves the stability of small molecule aminopolymers against oxidative degradation. In oxidizing conditions, small-molecule amines such as ethylenimines can undergo chain rearrangement reactions to form six-membered rings that quickly oxidize to piperazinones. Propylenimines, on the other hand, form four- or eight-membered rings, which are more stable toward deactivation. Further, PPI having a lower molecular weight of 700 g mol⁻¹ and 1000 g mol⁻¹ maintained about 65% of its amine efficiency. On the other hand, samples with higher molecular weight PPI (6700 g mol⁻¹ and 36 000 g mol⁻¹) retained over 83% efficiency in the presence of an oxidative environment.¹⁰⁸ This is due to the inaccessibility of oxygen to penetrate the polymer block to carry out the oxidative reaction. Sarazen et al. demonstrated the use of branched PPI along with acid initiators such as HBr, HClO₄, HCl, and CH₃SO₃H, where sorbents with HBr retained 76% of their CO₂ adsorption capacity after 12 h of oxidation in air at 110 °C.¹²⁹ Recently, Rosu et al.¹³¹ evaluated the effect of extended aging (∼2 years) of linear PPI and observed that the linear PPI with a molecular weight of 700 g mol⁻¹ retained 68% of its CO₂ adsorption capacity after oxidation for 12 h in 21% O₂/He stream under 400 ppm CO_2. Also, the same sample could retain 66% of its adsorption capacity during short TSA cycles (6 cycles, 1 h adsorption at 35 °C, and 30 min desorption at 110 °C).

Recently, Kumar et al.⁴¹ developed alkyl-amine rich small molecules modified with ethylenediamine/propane-1,3-diamine for the DAC process. The accelerated oxidation stability experiments at high temperatures (90 and 110 °C) under 21% O₂ for 24 h revealed that the ethylenediamine-modified sorbent retained 35% and 20% CO₂ adsorption capacity and 39% and 23% amine efficiency at 90 and 110 °C, respectively. On the other hand, propane-1,3-diamine impregnated sorbent maintained 22% and 13% CO₂ adsorption capacity and 20% and 10% amine efficiency at 90 and 110 °C, respectively. They concluded that the aromatic cores could change the distances between the oligomeric alkyl amine chains and hinder the electronics of the impregnated amine molecules and thus show lesser oxidative stability compared to the PPI-based sorbents.

3.3.3 Moisture stability of amine-based solid adsorbent for DAC. The performance of DAC technology largely depends on the climate at the given location, particularly the temperature and relative humidity. Typically, the water concentration present in the ambient air is about 50 to 75% times higher than the CO₂ concentration. For the practical use of an amine-based adsorbent in the DAC process, evaluating its hydrophilic/hydrophobic nature is necessary to understand the positive and negative effects of moisture/humidity on adsorption capacity. Although moisture often increases the CO₂ uptake of amine-modified adsorbents,^{39,41,119,123} the lack of stability of the material pore structure in the presence of water is considered one of the shortcomings in practical DAC application,⁷⁷ where low-quality steam is used for the regeneration of adsorbents.¹⁵⁵ Further, excess humidity can result in water condensation inside the pores and sometimes even washes the aminopolymer from the adsorbent, thus reducing their CO₂ uptake performance.^84,105 Therefore, the amine-based adsorbent needs to be structurally sturdy and should exhibit cyclic CO₂ adsorption performance under humid DAC conditions.

Steam or moisture-containing non-adsorbing gases, such as CO₂, can be employed as a purge gas for amine-based adsorbents, enabling pure CO₂ to be extracted.¹⁵⁶ Sayari et al. reported the mechanism behind the stability of amine-based adsorbents to humidity.⁷² Monoamine, triamine-grafted pore-expanded MCM-41 mesoporous silica (PE-MCM-41), and polyethylenimine impregnated PE-MCM-41 were studied to adsorb pure CO₂ and desorbed CO₂ in the presence of moisture (20 °C as dew point). It was found that when desorption was carried out under dry conditions (moderate to high temperature), the cycle stability dropped by 14–45% after 40 cycles due to the formation of urea as a byproduct or reaction between amine and CO₂. However, in the studied amine-based adsorbent, when desorbed in the presence of moisture, the cyclic stability was retained for up to 750 cycles without the formation of urea. The mechanism of urea formation was probed by ¹³C CP MAS NMR and DRIFT study. The MAS NMR peak at 164.6 ppm (for carbamate) and 1600 cm⁻¹ (NH₂ scissoring) was seen in amine-modified PE-MCM-41 when they underwent desorption under humid conditions. However, carbamate decomposes to produce urea in a dry desorption environment, which is shown by the new peak that appeared at 160.5 ppm in MAS NMR and two new bands (1658 and 1560 cm⁻¹) in the IR pattern as shown in Fig. 9.

Fig. 9 (a) ¹³C CP MAS NMR spectra, (b) DRIFT spectra for fresh urea, fresh mono-MCM-41 and after cyclic desorption in dry and humid conditions, and (c) reaction mechanism of amine and CO₂ to form urea as well as a carbamate. Reproduced with permission.⁷² Copyright 2010, American Chemical Society.

The author also reported that it is also possible to deactivate the strongly formed urea groups in the cyclic adsorbent via hydrolysis at 200 °C under a flow of nitrogen-containing as little as 0.15% RH (dew point at 20 °C) for 24 h. AEATPMS-modified silica gel and AEAPDMS-modified nano-fibrillated cellulose have been reported as one of the most stable adsorbents under thermal swing regeneration (in a humid environment), preventing the formation of urea.^112,113

Furthermore, MOFs in the DAC process require partial or complete drying of the gas streams. Some MOFs are hydrophilic and can end up adsorbing more H₂O than CO₂.¹⁵⁷ However, amine-modified MOFs prove to be more stable under humid conditions of the DAC process. In the case of PEI/MIL-101-Cr,¹³⁵ and N₂H₄/Mg₂(dobdc)¹¹⁸ the measured CO₂ adsorption capacity before and after the exposure to humidity of 32% RH and under 82% RH, respectively, showed little change in the adsorption capacity, suggesting the material is stable under DAC working conditions.

3.4 Adsorption thermodynamics of amine-based solid adsorbent for DAC

Adsorption thermodynamics describes the energetic interaction between adsorbate and the adsorbent's surface as a function of state variables (pressure, temperature, chemical composition, and others).¹⁵⁸ To design an adsorption-based system, it is essential to understand the thermodynamic properties, especially the enthalpy of adsorption.¹⁵⁹ The enthalpy of adsorption is defined as the heat released when adsorbate binds to an adsorbent. Experimentally, the heat (enthalpy) of adsorption (ΔH_ads) is evaluated by either direct or indirect methods.¹⁶⁰ In the direct method (usually called the calorimetric method), the heat released upon the contact between adsorbate and adsorbent is measured using a calorimeter (differential scanning calorimetry or microcalorimetry). In the indirect method, the enthalpy of adsorption is estimated from the adsorption isotherm data at multiple temperatures by fitting adsorption isotherms using Langmuir, Toth, or related thermodynamic models and then applying the Clausius–Clapeyron equation.¹⁶⁰ The enthalpy of adsorption can also be expressed as isosteric heat of adsorption (Q_st) at particular coverage. The isosteric heat of adsorption has the same magnitude as the enthalpy of adsorption but with an opposite sign (ΔH_ads = −Q_st).¹⁶¹ The isosteric heat of adsorption provides necessary information about the nature of the adsorption process (exothermic or endothermic) it allows quantifying the energy needed to desorb the adsorbate from the adsorbent.¹⁶² Among the two measurement methods, the isosteric heat measurement by the Clausius–Clapeyron equation is widely used compared to the calorimetric measurement.¹⁶⁰ However, the heat value calculated from the indirect method sometimes underestimates the actual heat value due to the inconsistency in adsorption fitting. Hence, for practical DAC application, the heat of desorption and the heat of regeneration, along with the heat of adsorption, are also calculated using calorimetry.¹⁶³

The isosteric heat of adsorption for CO₂ capture over a few reported amine-based adsorbents under DAC conditions comes under 49–130 kJ mol⁻¹ (Fig. 10, Table 5). These values depend on the amine–CO₂ interaction, type, basicity, and amount of amine loaded in adsorbent/amine support.¹⁶³ Among primary, secondary, and tertiary amines, the adsorbents loaded with primary amine exhibit high heat of CO₂ adsorption, even at very low CO₂ concentrations due to their aggressive interaction with CO₂.¹⁶³ For instance, the bare adsorbent such as silica exhibited heat of CO₂ adsorption typically in the range of 20 to 40 kJ mol⁻¹.¹⁶⁴ However, introducing primary amine such as 3-aminopropyl silane to the silica substrate greatly enhanced its heat of CO₂ adsorption in the range of 65 to 130 kJ mol⁻¹ depending on the density of amine sites.^163,165 With the increase in amine density in the substrate (>1.2 mmol N per g of adsorbent), the heat of CO₂ adsorption at zero coverage was observed to show an increasing trend, primarily due to the formation of strong alkylammonium carbamate species (92 kJ mol⁻¹) in dry conditions by allowing multiple amines to interact with one CO₂ molecule. Further, at a higher amine loading, the heat of adsorption at zero coverage increased significantly once the average amine spacing reached a value below ca. 11–12 Å, assuming all sites are equally spaced. On the other hand, at low amine density in the substrate, the heat of CO₂ adsorption was decreased to 40–60 kJ mol⁻¹ similar to bare SBA-15 support because of the less interaction of CO₂ with the fewer isolated amine species present in the substrate. In such a case, carbamic acid will be preferred over carbamate.^163,164 Similar observation was found by Potter et al.⁹¹ by revealing the heat of CO₂ adsorption (using microcalorimetry) of different species formed (using in situ FTIR) during the interaction of CO₂ and 3-aminopropyltriethoxysilane (APTES) grafted mesoporous alumina (D-Al₂O₃ and O-Al₂O₃). Lower APTES loading into alumina resulted in the formation of carbamic acid (due to more isolated amine species), which is generally less energetically stable and exhibits low heat of CO₂ adsorption (45–65 kJ mol⁻¹) in alumina pores as compared to alkylammonium carbamate (98 kJ mol⁻¹). Further, the same group also reported that the resultant heat of CO₂ adsorption of APTES modified alumina also depends upon the initial basicity of alumina surface. Alumina D-Al₂O₃ is more basic than O-Al₂O₃ due to presence of hydroxyl groups and O²⁻ group (heat of CO₂ adsorption: 79 kJ mol⁻¹vs. 52 kJ mol⁻¹) and it was believed that the heat of adsorption would further increase after amine loading on D-Al₂O₃. However, after the same equivalent APTES loading, the initial heat of CO₂ adsorption in D-Al₂O₃ was notably lower than O-Al₂O₃ due to the restriction in interaction of CO₂ with strong binding sites of alumina by the APTES grafting process. Therefore, D-Al₂O₃ overcame a larger energy barrier to interact with CO₂.⁹¹

Fig. 10 Heat of adsorption for CO₂ capture over literature-reported amine-based adsorbents (based on Table 5).

Table 5 Heat of adsorption of amine-functionalized solid adsorbents for DAC

Amine based composite	Heat of adsorption (kJ mol^−1)	Method	Ref.
a Isosteric heat of CO2 adsorption at zero coverage obtained by the Clausius–Clapeyron equation and isotherms fitted by the respective thermodynamic models like virial, Toth, etc. b Isosteric heat of CO2 adsorption at very low surface coverage obtained by the Clausius–Clapeyron equation using the experimental isotherm data.
Fumed silica + PEI-M	83	Calorimetry	81
Mesoporous silica + PEI	90	Calorimetry	124
SBA-15 + PPI	105^a	Toth isotherm	108
PEHA-PO-1-2-Sipernat 50S	72	Calorimetry	40
Mg2(dobpdc) + ed	49–51^a	Dual-site Langmuir–Freundlich	115
Mg2(dobdc) + hydrazine	118^a	Langmuir–Freundlich	118
Mg2(dobdc) + mmem	71^a	Dual-site Langmuir–Freundlich	42
MIL-101-Cr + PEI	70^b	Directly from isotherm data	135
Cr-MIL-101-SO3H + TAEA	87^a	Triple-site Langmuir	140
O-Al2O3 + APTES	98	Calorimetry	91
D-Al2O3 + APTES	69	Calorimetry	91
MCF + APTMS	130^b	Directly from isotherm data	84
MCF + MAPS	88^b	Directly from isotherm data	84
Nano fibrillated cellulose + APDES	73^b	Directly from isotherm data	119
SBA-15 + APTMS	65^a	Langmuir	138
Pore-expanded MCM-41 + TRI	92^b	Directly from isotherm data	33
PPN-6 + DETA	54^a	Triple-site Langmuir	120
Hybrid silica + APTES	85 ± 5	Calorimetry	111

---

Adsorbent modified with secondary amine (associated with methylaminopropyl silane, MAPS) exhibited heat of CO₂ adsorption similar to that of primary amine at a similar amine density but is less effective for CO₂ capture from air.^84,163 For example, Didas et al.⁸⁴ reported the high heat of CO₂ adsorption (130 kJ mol⁻¹) for APTMS grafted mesoporous cellular form (MCF) as compared to those grafted with MAPS (88 kJ mol⁻¹). Notably, tertiary amines such as dimethylaminopropyl silane (DMAPS) are inefficient for DAC application due to their poor interaction with CO₂. On the other hand, if an adsorbent is loaded with amines of all types, including primary, secondary, and tertiary, the net heat of CO₂ adsorption of the adsorbent is not equal to the sum of the heat of CO₂ adsorption of individual amine species of the adsorbent, but rather depends on the individual content of amines and their reactivity toward CO₂. For instance, commercially available low-molecular-weight branched PEI has a carbon : nitrogen molar ratio of 2 : 1 [(C₂H₅N)_n, linear form] with a distribution of amine sites as 44% primary, 33% secondary, and 22% tertiary amines.¹⁶³ Primary, secondary, and tertiary amines exhibit typically heat of CO₂ adsorption values around 84 kJ mol⁻¹, 72 kJ mol⁻¹, and 48 kJ mol⁻¹, respectively.^34,166 Therefore, the heat of CO₂ adsorption of the highest PEI-loaded adsorbent (PEI-7.5/SBA-15) is found to be 93 kJ mol⁻¹, which is less than the summation of the heat of reaction of its stand-alone constituent amines. This observation can be attributed primarily to the fact that primary, and a few secondary amines present in PEI are majorly responsible for the adsorption of CO₂ even at a low concentration of CO₂. It is demonstrated that PEI-SBA-15 exhibited optimal heat of CO₂ adsorption with better CO₂ uptake despite the low concentration of CO₂, and hence is perfect for DAC applications.¹⁶³ In a similar work, Pang et al.¹⁰⁸ reported that compared to PEI–silica composite (50–80 kJ mol⁻¹), the isosteric heat of CO₂ adsorption at zero surface coverage on PPI–silica composite (105 kJ mol⁻¹) is considerably greater, presumably due to the higher basicity of propylenimines than ethylenimines, which resulted in strong adsorption of CO₂ on PPI.

Although high enthalpy of adsorption for DAC is the prerequisite criteria for CO₂ capture from air, it imposes a high parasitic energy load on the plant. Therefore, researchers focus on incorporating smaller chain amine into adsorbents without compromising the CO₂ adsorption capacity.^40,167 Secondary amino groups have been reported to bind CO₂ less strongly than primary amino groups.³⁰ This is advantageous during the adsorbent's endothermic regeneration step when CO₂ has to be released. Therefore, the amount of energy required for the desorption could be lowered by employing polyamines mainly containing secondary amino groups. Goeppert et al.⁸¹ estimated the heat of CO₂ adsorption, the heat of CO₂ desorption, and the heat of CO₂ regeneration respectively as 83 kJ mol⁻¹ (at 85 °C), 82 kJ mol⁻¹, and 170 kJ mol⁻¹ (from 25 to 110 °C) for branched PEI based fumed silica using calorimetry. On the contrary, when the fumed silica is modified with linear PEI, which consists of secondary amine, the associated heat of CO₂ adsorption (42 to 53 kJ mol⁻¹) and desorption (40 to 49 kJ mol⁻¹) decreased significantly.¹⁵⁰ The estimated heat of regeneration is almost half of the heat of regeneration when considering a 20% aqueous solution of MEA (330 kJ mol⁻¹, considering an absorption of 0.4 mol of CO₂ for 1 mol of MEA) and is therefore energetically favorable. To further reduce the associated thermal energy for CO₂ regeneration without compromising its uptake, the same group, Goeppert et al.⁴⁰ estimated the heat of CO₂ adsorption of polyamines (PEHA and TEPA) supported Sipernat 50S silica with the addition of stabilizing agent propylene oxide (PO) on direct isothermal calorimetry. With the addition of PO, the heat of CO₂ reaction for PEHA-PO-1-2/50S was decreased to 72 kJ mol⁻¹, which is 10 kJ mol⁻¹ less than PEHA/50S.

Considering humid conditions, although the CO₂ adsorption capacity of amine-based adsorbents rises, especially at ultra-diluted CO₂ pressures, the H₂O content must be kept as low as possible.¹⁶⁸ For DAC, temperature swing desorption or temperature-vacuum swing desorption is preferentially used to recover the adsorbents during the repetitive adsorption–desorption cycles. High moisture sorption in the adsorbent leads to higher parasitic energy requirements during the regeneration of a sorbent. It has been observed that, during the regeneration of adsorbent under TVSA, if the H₂O loading in the air increases from 20 to 60 RH% at 30 °C, the thermal energy demand increased from 532 to 586 kJ mol⁻¹ CO₂.¹⁶⁸ This is because the total energy required to regenerate an adsorbent during the co-adsorption of CO₂ and H₂O in the DAC condition is the sum of the sensible heat required to bring the material to desorption temperature and the heat input for the desorption enthalpy of H₂O and CO₂. Typically, during CO₂ adsorption under dry conditions, the enthalpy change of reaction values were reported as 50 to 118 kJ mol⁻¹ and 56 kJ mol⁻¹ for hydrous conditions, additionally with 47 to 53 kJ mol⁻¹ for H₂O adsorption.^34,113 Since moisture in the air is inevitable, the net thermal regeneration energy for any practical DAC process could be decreased significantly by increasing the CO₂ adsorption capacity of the material beyond 2 mmol g⁻¹ to nullify the sensible heat effect.¹⁶⁸

3.5 CO₂ over other gases selectivity of amine-based solid adsorbents for DAC

The high selective adsorption of CO₂ from the other bulk components of the air (78% N₂, 21% O₂, and 1–2% H₂O) is the most essential and desirable criteria to be achieved for any amine-based adsorbent. Usually, for DAC the selectivity for CO₂ is calculated based on the ratio of CO₂ uptake capacity (at 0.4 mbar) to that of the other components (for example, N₂ or O₂) according to their partial pressure in the air.¹⁶⁹ The adsorbent having high selectivity for CO₂ has three crucial significances in DAC, including (i) production of ultra-high pure (UHP) CO₂, (ii) reduction of cost associated with the CO₂ capture process, and (iii) requirement of less energy for transportation and compression. Since air contains a large concentration of N₂ and O₂, it is crucial to develop amine-based adsorbents that ideally would not react with either N₂ or O₂ at room temperature and atmospheric pressure. Further, the CO₂ purity should be greater than 90% to meet the department of energy (DOE) demand to be utilized for further value-added applications such as underground storage.¹³⁸

Belmabkhout et al.^33,143 reported a novel adsorbent (triamine-grafted pore-expanded mesoporous silica TRI-PE-MCM-41) that exhibited excellent CO₂/N₂ and CO₂/O₂ selectivity (∼ infinity) with high CO₂ uptake (0.9 ± 0.09 mmol g⁻¹) when exposed to the simulated air containing CO₂ (0.03%) and N₂ (79.98%) and O₂ (19.99%) at 25 °C and 1 bar. The experiment was conducted in a breakthrough setup, where both O₂ and N₂ appeared in the column downstream almost immediately after the process started, indicative of a minimal adsorption capacity for N₂ and O₂. Further, the same group also reported the adsorption performance of TRI-PE-MCM-41 in the presence of simulated air containing CO₂ : N₂ = 0.04% : 99.96% in a dry condition as well as in presence of moisture 27% and 64% relative humidity (RH), at 25 °C and 1 bar.¹⁴³ TRI-PE-MCM-41 excelled in high CO₂–amine interaction as indicated by a steep rise in a breakthrough curve, with high CO₂ adsorption capacity and H₂O uptake of 4.7 mmol g⁻¹ and 6.29 mmol g⁻¹ at 27 and 64% RH respectively, due to the partial formation of bicarbonate by the reaction of amine molecules with CO₂.¹⁴³ The corresponding CO₂ adsorption capacity was 32 and 56% higher than the capacity under dry conditions. Advantageously, even in the presence of moisture, TRI-PE-MCM-41 exhibited high CO₂/N₂ comparable to that observed under dry conditions (Fig. 11).

Fig. 11 (a) Column-breakthrough curves for CO₂ : carbon-free air = 0.03 : 99.97% mixture at 25 °C and 1 bar in dry condition. (b) Breakthrough curves for CO₂ : N₂ = 0.04 : 99.96 mixture at 25 °C and 1 bar with and without the presence of moisture (27% and 64% RH). Reproduced with permission.³³ Copyright 2009, ACS.

Lu et al.¹²⁰ investigated the CO₂ adsorption performance of ethylenediamine and diethylenetriamine (DETA) modified porous polymer networks PPN-6-CH₂DETA in the condition similar to air, considering the partial pressure of the gases present in the air. PPN-6-CH₂DETA exhibited an extraordinarily high CO₂/N₂ selectivity (3.8 × 10¹⁰), CO₂/(N₂ + O₂) selectivity (3.6 × 10¹⁰) as calculated using ideal adsorption solution theory (IAST), CO₂ adsorption capacity (1.04 mmol g⁻¹), high stability towards moisture, and excellent CO₂ purity (99.999%). Compared to ethylenediamine (which has only short diamine chains), DETA which has also a third amine available for interaction with the CO₂ molecule can be loaded at a higher loading into PPN-6 without any pore blockage. The high loading capacity and chemical interactions with CO₂ boost the CO₂ selectivity of DETA by 7 orders of magnitude than the corresponding ethylenediamine-modified PPN-6-CH₂. The group also marked that O₂ has minimal interference in competing for adsorption sites against CO₂ when PPN-6-CH₂DETA is subjected to air due to its strong interaction with CO₂–DETA, highlighting the suitability of PPN-6-CH₂DETA for DAC application.

Similarly, Lee et al.¹¹⁵ reported the CO₂/N₂ selectivity of 70 000 and CO₂ purity of 97% in the simulated air [390 ppm CO₂ and 21% O₂ balanced with N₂, and Ar (99.999%)] for en-Mg₂(dobpdc) MOF (en: ethylenediamine), which is higher than mmen-Mg₂(dobpdc), (mmen; N,N-dimethylethylenediamine) (corresponding selectivity 49 000 and purity 96%).⁴² The high CO₂/N₂ selectivity was attributed to carbamate species (–NHCOOH) formation by the specific interactions between the amine groups of en and CO₂. Apart from the preferential interaction of CO₂ with amine-based adsorbent, CO₂ purity also depends upon the desorption process cycle in the practical application. Compared to TSA, TVSA would be sufficient for CO₂ purification. The vacuum degasses the amine-based adsorbent more suitably, ensuring minimum loss of CO₂ at a shorter cycle time. The analogous observation was reported by Stuckert et al.,¹³⁸ where SBA-15-APTMS exhibited ∼99% CO₂ purity under TVSA (0.01 bar and 95 °C) conditions as compared to 9% CO₂ purity observed in TSA (1 bar, and 120 °C) under dry DAC condition (dry ambient air containing 395 ppm CO₂).

3.6 Shaping and structuring of amine-based solid adsorbent for DAC

Amine-based adsorbents for carbon capture can satisfy CO₂ capture needs while consuming the least energy. However, this results in a significant pressure drop and issues related to the mass transfer and attrition loss,¹⁷⁰ particularly when the adsorbents are used in powdered form in a cyclic adsorption process, including PSA and/or TSA. The use of structured monolithic adsorbents has been proposed to address the drawbacks of randomly packed beds, thereby increasing the energy efficiency and economic value of the carbon capture process.¹⁷¹ Recently, additive manufacturing, commonly known as 3D printing, has shown to be a viable platform for shaping porous materials and creating scalable configurations that address these concerns.¹⁷² Honeycomb monolith structure has been vastly used for air drying and CO₂ capture and separation application,¹⁷³ as the honeycomb monolith structure has the advantage of low-pressure drop at a high flow rate as well as superior heat and mass transfer compared to the packed bed.¹⁷⁴ Typically, mineral-like cordierite is commonly used as construction material for honeycomb monolith synthesis, due to its low thermal expansion. Such monoliths can be wash-coated with high surface area porous material and coated with amine for CO₂ capture applications.¹⁷⁵ Kulkarni and Sholl utilized a large structured monolith contractor made of amine-functionalized silica adsorbent (TRI-PE-MCM-41) for TSA-based DAC application, using the standard shipping container-sized air capture unit, they could get a daily throughput of ∼1.1 ton CO₂ at 85.5% purity.¹⁵⁵ Sakwa-Novak et al.¹⁷⁶ evaluated the material characteristics and performance of alumina-supported PEI sorbents in their powdered and honeycomb monolith forms. Powder sorbents made by grinding unfunctionalized monolith and then functionalizing with PEI were compared to the monolithic alumina modified with PEI. The monolithic adsorbents showed a CO₂ adsorption capacity of 0.75 mmol g⁻¹ compared to the powdered adsorbents (0.7 mmol g⁻¹) at 30 wt% amine loading. The slightly higher adsorption capacity was attributed to the high amine loading (higher volume of PEI content) of monolithic sorbents, implying that some PEI was deposited in the macropores of the structure (Fig. 12). After being exposed to running steam for varied durations, the CO₂ adsorption capabilities of PEI-impregnated extruded alumina monolith was stable over five cycles of CO₂ adsorption and has shown a volumetric capacity of 350 mol CO₂ per m³ with a faster (350 min) equilibration time under 400 ppm CO₂/N₂, 30 °C, and 0.4 m s⁻¹ gas flow conditions. On the other hand, Wijesiri et al.¹⁷⁷ evaluated a PEI impregnated mesocellular foam silica pellet for capturing CO₂ from 420 ppm CO₂ in N₂ under both dry and humid conditions. At a temperature of 46 °C, the maximum capacity of 2.52 mmol g⁻¹ was achieved at a moisture content of 2% mol H₂O and 1.94 mmol g⁻¹ under dry conditions.

Fig. 12 Alumina-supported honeycomb monolith and SEM image of its square channel cross-section, impregnated with PEI for DAC application, adapted with permission.¹⁷⁶ Copyright 2016, WILEY-VCH Verlag GmbH & Co.

3.7 Moisture swing sorption: a new emerging CO₂ regeneration technology for DAC

K. S. Lackner¹⁷⁸ came up with the distinctive concept of moisture swing sorption for DAC to limit the exorbitant expense of amine-based chemisorbent regeneration. The idea of moisture swing adsorption is very similar to green leaves, which capture net CO₂ in the presence of sunlight and release it at night. Moisture swing CO₂ sorbent act as an “artificial leaf” to bind CO₂ in nearly dry conditions and release it in humid environments by changing the moisture content of the surrounding air. Therefore, the interconversion of carbonate and bicarbonate moieties proposes the adsorption and desorption process.¹⁷⁹ Due to their minimal environmental and economic cost by avoiding expensive heating/cooling cycles¹⁸⁰ or significant changes in the pressure over the sorbent, these materials are promising for reversible CO₂ capture through humidity swing and appealing for large-scale applications.¹⁸¹ The fact that the humidity swing process needs to operate in a dry area, and the limited adsorption capacity are the few potential downsides of a humidity swing method. Further, a large-scale DAC plant based on moisture swing requires significant water to operate. It is estimated that about 10 gigatons of CO₂ could use as much as 100 km³ of water,⁹ compared to a world water consumption of 3600 km³.¹⁸² The reaction pathway of CO₂ sorption/desorption on a moisture-swing absorbent reported by Shi et al.¹⁸¹ is shown in Fig. 13 and eqn (7)–(10).

H₂O ⇌ H⁺ + OH⁻ (7)

CO₃²⁻ + H⁺ ⇌ HCO₃⁻ (8)

OH⁻ + CO₂ ⇌ HCO₃⁻ (9)

HCO₃⁻ + HCO₃⁻ ⇌ CO₃²⁻ + CO₂ + H₂O (10)

CO₃²⁻·nH₂O ⇌ HCO₃⁻·m₁H₂O + OH⁻·m₂H₂O + (n − m₁ − m₂ − 1)H₂O (11)

Fig. 13 Reaction pathway of CO₂ absorption/desorption on a moisture-swing absorbent. The empty-fresh state is the sorbent in dry condition with a few water molecules in the surrounding. The empty-dry state is when H₂O splits into H⁺ ion and OH⁻ ion, which is ready to absorb CO₂, and H⁺ ion is combined with CO₃²⁻ forming HCO₃⁻ ion [eqn (7) and (8)]. The full-dry state is the fully loaded sorbent in the dry condition [eqn (9)], eqn (7)–(9) present the absorption process. The empty-wet state is the regenerating absorbent releases CO₂ in the wet condition (desorption [eqn (10)]) reproduced with permission.¹⁸¹ Copyright 2016, Wiley.

Typically, ion hydration/dehydration mechanism in ion-exchange resin plays a greater role in the moisture-swing sorption system. The eqn (11) belongs to the hydration of water in the moisture-swing adsorbent, representing the equilibrium between carbonate and bicarbonate ions. The lower is water molecules (n) in dry conditions, the higher is the OH⁻ ions, and hydrated bicarbonate generation, resulting in larger CO₂ capture in the sorbent. However, at higher water molecules in wet conditions, carbonate ion hydration occurs, resulting increase in hydration Gibbs free energy and the release of CO₂. As evidenced from eqn (11), Shi et al.¹⁸¹ reported that with an increase in 8 to 60 water molecules, the difference in hydration Gibbs free energy increases quickly from a negative to a positive value, then becomes stable at a plateau of 15 kcal mol⁻¹ in bulk water. Many moisture-swing sorbents consist of quaternary ammonium ions (NR₄⁺), which are strong bases in aqueous solutions similar to Na⁺ ions. The cations in the matrix or support are replaced by anions such as carbonate, bicarbonate, and hydroxide ions during CO₂ adsorption. Their relative abundance depends on the CO₂ loading of the resin and the quantity of moisture present.⁹ For moisture swing adsorption, the reference material is Excellion membrane (I-200), a commercially available extruded polyethylene membrane containing crosslinked chloromethylated polystyrene resin with quaternary ammonium hydroxide groups (NMe₃⁺–MS OH⁻).¹⁷⁹ This membrane has been utilized to capture CO₂ due to humidity changes. However, the Excellion membrane displayed slow desorption and absorption profiles with an overall rate of 2.1 × 10⁻³ mmol g⁻¹ min⁻¹ and a moderate swing size of 1.3 × 10⁻¹ mmol g⁻¹ (Table 6, entry 1).¹⁷⁹ Further, the Excellion membrane required over 2 h to capture and release 0.13 mmol of CO₂ per 1 g of material to complete a cycle. To overcome the slow rate of adsorption/desorption, new materials were explored where ammonium ions lacked abstractable β-hydrogens, and the material did not contain ester or amide bonds that could be hydrolysed.¹⁷⁹ He et al.¹⁷⁹ designed a templated p(NMe₃⁺–MS OH⁻) carbon black sorbent containing three methyl substituents and one benzyl substituent which links the ammonium group to the polymer backbone. The material outperformed the other four materials (Table 6). The overall rate of CO₂ capture and release for CB-g-p(NMe₃⁺–MS OH⁻) was found to be 3 times higher than the Excellion membrane. Moreover, the swing size was 3.8 times larger owing to its surface area of 6.2 m² g⁻¹ and a highly porous structure with an average pore size of 0.83 mm. On the other hand, compared to the materials which were made by conventional high internal phase emulsion (HIPE) polymerization, the increased pore size found in porous polymers generated using Pickering HIPE polymerization improved CO₂ capture ability.¹⁷⁹ The biggest swing size was 7.2 × 10 ⁻¹ mmol g⁻¹ (Table 6, entry 6) and the quickest overall rate was 2.5 × 10⁻² mmol g⁻¹ (Table 6, entry 4), both of which were 5.5 and 11.9 times higher than those achieved with Excellion membrane.¹⁸⁴ The ideal parameters of humidity swing sorbents for traditional point-source capture and those for air capture were found to be significantly different. Micro-/mesoporous materials with a high surface area are preferred for point source capture, while larger pore sizes and better interconnectivity pores (as seen in colloidal crystal templated material and the HIPE-based material) are better for direct air CO₂ capture using a humidity swing process.¹⁷⁹ Further, considering the energy consumption, the moisture swing adsorbents need much less energy for CO₂ desorption (∼50 kJ mol⁻¹)¹⁷⁸ compared to other conventional adsorbents, which makes it economical for large-scale CO₂ capture from the air and further utilization. Notably, the free energy released by water evaporation is utilized in moisture-induced cycle while avoiding the use of heat for sorbent regeneration.¹⁸³

Table 6 Moisture swing CO₂ capture by absorbents for DAC

Material	CO2 concentration (ppm)	Relative humidity (RH%) cycle	Swing size or CO2 sorption capacity (mmol g^−1)	Absorption rate (mmol g^−1 min^−1)	Desorption rate (mmol g^−1 min^−1)	Overall rate^b (mmol g^−1 min^−1)	Ref.
a Adsorption half time (minute). b Overall rate = 1/{(1/absorption rate) + (1/desorption rate)}.
Excellion membrane	Pure CO2 (90%)	20%	0.13	4.0 × 10^−3	4.5 × 10^−3	2.1 × 10^−3	179
CB-g-p(NMe3^+–MS OH^−)	400	20% to 95%	0.14	1.8 × 10^−2	1.2 × 10^−2	7.2 × 10^−3	179
Colloidal crystal templated p(NMe3^+–MS OH^−)	400	20% to 95%	0.37	2.8 × 10^−2	1.8 × 10^−2	1.1 × 10^−2	179
HIPE templated p(NMe3^+–MS OH^−)	400	20% to 95%	0.49	1.1 × 10^−1	3.3 × 10^−2	2.5 × 10^−2	179
PolyHIPE (1.5% crosslinking)	400	20% to 95%	0.23	3.2 × 10^−2	3.0 × 10^−2	1.5 × 10^−2	184
Pickering HIPE polymer	400	20% to 95%	0.72	3.7 × 10^−2	2.7 × 10^−2	1.6 × 10^−2	184
CB-g-xPCMS-OH	400	20% to 95%	0.14	2.0 × 10^−2	1.5 × 10^−2	8.6 × 10^−3	185
PS-CC-film/VBTMACl/PMVPMACl/M/20%	400	20% to 95%	0.57	6.4 × 10^−3	1.4 × 10^−2	4.4 × 10^−3	186
Polypropylene and a resin – composite sheet comprising quaternary ammonium functional groups	400	0.5%	0.82	—	—		187
Q-cellulose	400	20% to 90%	0.18	9.8 (25 °C)^a			188
(QCS)/poly(vinyl alcohol)-hybrid aerogels	400	3% to 95%	0.18	10.72 (20 °C)^a			189

---

3.8 Techno-economic analysis of amine-based adsorbents for DAC

The techno-economic analysis is carried out to predict the economic viability of the DAC technology using the energy and cost estimation of the overall process.¹⁵ In this review, the techno-economic analysis of DAC based on amine-based adsorbents is briefly discussed. Besides the adsorbent's properties, the cost of CO₂ capture using amine-based adsorbents for DAC is highly dependent on the design of the adsorbent contactor and the regeneration process.^15,190 Since a high airflow rate over the adsorbent contractor is required for DAC, air compression may not be a good alternative, leaving temperature and vacuum the only viable regeneration drivers. As a result, the vacuum-thermal swing adsorption (TVSA) cycle is used in most sorbent-based DAC systems.³⁶ DAC companies such as Climeworks and Global Thermostat, use the contractor of air-filter-like structures and honeycomb monoliths, respectively, and a TVSA process for CO₂ capture from air.³⁶

Sabatino et al.³⁶ recently reported a sorbent-based DAC system using the TVSA process, where the process comprises (i) adsorption, (ii) purge, (iii) regeneration, and (iv) repressurization steps. The adsorption step was carried out at ambient conditions (T_ads = 20 °C, CO₂ concentration: 400 ppm), and the preheating and regeneration step (T_des = 100–120 °C) were carried out with the virtue of simultaneous heating and vacuum (0.1 bar) to remove N₂ from void space and recover the captured CO₂ from the adsorbent respectively. The group estimated the energy, and cost of DAC using the VTSA process on four amine-based adsorbents namely APDES–NFC, TRI-PE-MCM-41, MIL-101(Cr)-PEI-800, and Lewatit VP OC 106. In all cases, the CO₂ purity was found to be in the range of 94% to 99%, with water being the main impurity. The overall CO₂ capture cost was estimated to be 149–427 $ per tCO₂. For these four amine-based sorbent processes, the total energy demand varied in a range of 4.9–13.3 MJ kg⁻¹ CO₂, where the thermal energy had the most significant contribution (4–11.8 MJ kg⁻¹), whereas the electrical energy accounted for 0.8–1.8 MJ kg⁻¹, which indicated that the results were in line with the Climeworks and Global Thermostat report.^35,191,192 The process developed by Climeworks uses special cellulose fiber supported by amines in a solid form requiring 0.7–1.0 MJ kg⁻¹ (electrical) and 5.4–7.2 MJ kg⁻¹ (thermal) energy.³⁵ An entire DAC cycle of Climeworks runs about 4–6 h and produces 99.9% pure CO₂, using waste heat as a thermal source for regeneration.³⁵ Global Thermostat, a USA-based company, used amine-polymer adsorbent for a TVSA-based DAC system. The adsorption is carried out under ambient conditions, and the desorption is accompanied by vacuum and saturated steam at 85–95 °C, producing CO₂ purity >98.5%. The total cycle time of the overall process is below 30 min, and the net thermal and electrical energy demands were 4.2–5.0 MJ kg⁻¹, and 0.54–0.94 MJ kg⁻¹, respectively.¹⁹³

A similar observation was also found by Sinha et al.¹⁹⁴ using MIL-101(Cr)-PEI-800 and mmen-Mg₂(dobpdc) amine append MOFs for TVSA run DAC system, where T_ads = 20 °C, CO₂ concentration: 400 ppm, T_des = 100 °C at 1 atm pressure. A study showed that sorbent would require a thermal energy demand of about 5.11 MJ kg⁻¹ for the MIL-101(Cr)-PEI-800 and 3.6 MJ kg⁻¹ for the mmen-Mg₂(dobpdc). Their work showed that the net cost of CO₂ removal would be in the range of 75–140 $ per tCO₂ and 60–190 $ per tCO₂ for the former and later sorbents respectively, considering the lifetime of adsorbent between 1 and 3 years. However, the CO₂ capture cost of MIL-101(Cr)-PEI-800 and mmen-Mg₂(dobpdc) using the same TVSA process could be reduced to below 75 $ per tCO₂ and 82 $ per tCO₂ if the sorbent cost was lowered to 15 $ per kg and 50 $ per kg respectively.¹⁹⁵ Sadiq et al.¹⁹⁶ also demonstrated a pilot-scale version of a mobile DAC machine using MOF–polymer nanocomposite using the TVSA process. The adsorption was carried out at ambient conditions by flowing the air over the adsorbent bed at a high flow rate (50 m³ h⁻¹) with the help of a blower. The adsorbent was coated as a thin layer onto long resistive heating sheets. The desorption was done by heating the adsorbent bed to 80 °C under a vacuum (0.037 bar). The plant exhibited the lowest energy consumption for regeneration (5.76 MJ kg⁻¹), moderate CO₂ purity (70–80%), and an operational cost of 35–350 $ per tCO₂.

In the same context, Kulkarni and Sholl used the TRI-PE-MCM-41 monolith and evaluated the energy and cost of the TSA-based DAC unit. The net energy demand was 0.79 MJ kg⁻¹ (electrical energy) and 6.0 MJ kg⁻¹ (thermal energy). Advantageously, the cost of capture was estimated to be ∼100 $ per tCO₂.¹⁵⁵ Bajamundi et al.¹⁹⁷ reported a TVSA-based DAC 8-bed system using amine-functionalized polystyrene. The resultant thermal energy demand was 27.4 MJ kg⁻¹ and the electrical energy demand was ∼5.0–26.3 MJ kg⁻¹. The DAC module was able to produce CO₂ purity up to 95%. Adsorption took place at ambient conditions (10–30 °C) and desorption was accomplished by heating the bed to around 80 °C coupled with a vacuum (<0.15 bar). Finally, McQueen et al.⁷⁹ also evaluated TVSA-based DAC using monolith contactors made of amine-modified adsorbents. The design (using the 5-step TVSA process) and calculation are similar to Sinha et al.,¹⁹⁴ but the cost of CO₂ capture is evaluated based on the regeneration using different modes of renewable and conventional energy sources. According to their research, the cost of CO₂ capture for DAC systems powered by geothermal, natural gas, and nuclear power plants was 205 $ per tCO₂, 223 $ per tCO₂, and 233 $ per tCO₂, respectively. The energy and cost associated with the DAC system of various amine-based adsorbents are listed in Table 7.

Table 7 Techno-economic details of amine-based adsorbents for DAC

Sorbent	CO2 conc. (ppm)	Operating conditions			Energy demand (MJ kg^−1)		Regen. source	CO2 purity (%)	CO2 production	DAC cost $ per tCO2	Ref.
		Adsorption	Desorption
		T ads (°C)	T des (°C)	P des (bar)	E electrical	E themal
T ads, adsorption temperature; Tdes, desorption temperature; Pdes, desorption pressure; HT, high temperature; Elec., electrical.
Amine-based cellulose (Climeworks)	400	Ambient	100	0.2	0.7–1.0	5.4–7.2	Waste heat	99.9	0.135 t CO2 per day	75	35, 192, 197
Amino-polymer (Global Thermostat)	400	Ambient	85–95	0.5–0.9	0.5–0.9	4.2–5.0	Steam	>98.5	—	113	193
Amine-polystyrene	400	Ambient	80	<0.15	5.0–26.3	27.4	Heating fluid	>95	3.4 kg CO2 per cycle	—	197
PE-MCM-41 + TRI	400	Ambient	110	1.4	0.78	6	Steam	88	1.1 tCO2 per day	∼100	155
PE-MCM-41 + TRI	400	Ambient	100–120	0.1	0.8–1.0	7.1–7.7	Heating fluid	>94	3.8–10.6 kg m^−3 h^−1	149–427	36
NFC + APDES	400	Ambient	100–120	0.1	1.1–1.3	8.6–10.1	Heating fluid	>94	3.8–10.6 kg m^−3 h^−1	149–427	36
Lewatit	400	Ambient	100–120	0.1	0.6–1.3	10.1–11.8	Heating fluid	>94	3.8–10.6 kg m^−3 h^−1	149–427	36
Mesoporous silica + PEI	395.5	Ambient	130	1	3.4	3.2	Elec.		40 tCO2 per day	108–152	124
MIL-101(Cr) + PEI-800	400	Ambient	135–480	1	5.1		HT steam	—	—	75–140	194
mmen-Mg2(dobpdc)	400	Ambient	135–480	1	3.6		HT stream	—	—	60–190	194
MOF–polymer	400	Ambient	80	0.037	2.4	5.7	Heating sheets	>70	6 kg CO2 per day	35–350	196

---

3.9 Comparison of evaluation parameters of amine-based solid adsorbent for DAC: a perspective

Since the existing CO₂ market is too small compared to the net CO₂ emitted every year (250 million tons per year vs. over 40 gigatons),¹⁹⁸ the installation of large numbers of DAC plants at affordable prices is gaining interest globally. In such a case, the specific focus must be on the material design, particularly on the amine-based solid adsorbent for DAC. It is also required to compare all of their major evaluation parameters on a single platform. Patel et al.¹⁹⁸ reported a critical assessment of evaluation parameters of the CO₂ adsorbent materials such as CO₂ adsorption capacity, CO₂/other gas selectivity, CO₂ kinetics, isosteric heat of adsorption, cost (adsorbent cost and CO₂ capture cost) and economic feasibility, and regenerability or cyclic stability of adsorbents along with their targeted limit. Although the evaluation parameters were initially designed for CO₂ capture from a point source, the same evaluation parameters could also be helpful for screening adsorbents for DAC with some modifications, as mentioned in Table 8.

Table 8 Benchmark target for an effective CO₂ sorbent (based on the data from Patel et al.)¹⁹⁸

Checkpoints	Evaluation criteria	Preferable desired limit
1	CO2 adsorption capacity	>2 mmol of CO2 per g of sorbent
2	CO2/other gas selectivity	≥100
3	CO2 adsorption kinetics	80% uptake within 2 min (rate: 1 mmol g^−1 min^−1)
4	Heat of adsorption	35–50 kJ mol^−1
5	Cyclic stability	>1000 cycles
6	Regeneration temperature	80–120 °C
7	Thermal and mechanical stability	Adsorbent should be stable at >150 °C, with high mechanical strength and low attrition index (AI)
8	Sorbent cost	<$10 per kg

---

However, while considering the amine-based adsorbents, none of them could satisfy all of the desired limits, making it challenging to design an optimum adsorbent for CO₂ capture. Considering few of the typical evaluation parameters (CO₂ uptake, heat of adsorption, and stability) of amine-based adsorbents akin to DAC conditions are compared and shown in Fig. 14. Though Mg₂dobpdc + ED and Mg₂dobdc + mmen appeared promising for DAC applications (Fig. 14) due of their high CO₂ uptake (>2 mmol g⁻¹), prolonged thermo-cyclic stability (∼100%), and marginal heat of adsorption (50–70 kJ mol⁻¹), rigorous process modeling and adsorbent evaluation metrics are required to screen the adsorbents for practical DAC application.^199,200 Nevertheless, a few of the amine-based adsorbents, such as MIL-101(Cr) + PEI, NFC + APDES, TRI-PE-MCM-41, and mmen-Mg₂(dobpdc) have already drawn the attention of numerous researchers, and few of them have already been employed in commercial or pilot scale TVSA based DAC processes, as also discussed in the previous section (refer section 3.8).

Fig. 14 Comparison of evaluation parameters of various amine-based adsorbents (data was chosen from Tables 2, 3 and 5).

In actual practice, the utilization of MOFs in DAC is limited due to their limited bulk-scale production and high capital cost. Despite the vast potential, the high cost associated with the precursor chemicals and machinery for the production of MOFs in large quantities retards its viability at present.²⁰¹ However, the use of MOF on polymer support or MOF–polymer nanocomposite could decrease the overall cost of the adsorbent.¹⁹⁴ The porous polymer support with low specific heat would be beneficial in transferring heat away from the material, thereby aiding in desorption.³⁴ Satyapal et al.³⁴ developed an HSC⁺ supported pressure swing adsorption (PSA) machine to remove CO₂ in space shuttle applications. The device was able to run hundreds of cycles and capture CO₂ up to 4% from dry air at ambient pressure with regeneration at a 1.33 mbar vacuum. After all, the field of DAC utilizing amine-based adsorbents is still in its infancy, and there is a lot of room for improvement in terms of sorbents and processes, of which only a handful are highlighted here.

(i) From the synthesis point of view, the focus should be on the cost-effective, facile, and faster synthesis of amine-grafted adsorbents, which typically consume several hours for amine condensation reaction.⁸³ Further, limited research articles are available on type 3 adsorbents and type 4 adsorbents (double-functionalized amine-based adsorbents) for DAC applications.

(ii) Since thermal swing or vacuum-thermal swing regeneration is suitable for DAC, especially for amine-based adsorbents, the adsorbent design mainly influences the thermodynamics and kinetics of CO₂ adsorption. The focus needs to be devoted towards developing large-scale structured adsorbents using emerging technology such as 3D printing to decrease the pressure drop along the adsorption column and better CO₂ kinetics without affecting the CO₂ adsorption performance as compared to the powdered form.

(iii) For a commercial DAC plant, the amine-based adsorbents must be stable for thousands of cycles. Though amine-impregnated adsorbents are suitable for high amine loading and hence excellent CO₂ uptake, they faced severe issues in cyclic adsorption–desorption stability due to amine leaching. Therefore, amine-grafted adsorbents could be the preferred option for DAC due to their high thermocyclic, moisture, and oxidative stability compared to the amine-impregnated adsorbents. However, a greater focus still needs to be devoted to increasing the CO₂ adsorption capacity of amine-grafted adsorbents.

(iv) The CO₂ desorption process needs to be less energy-intensive and low-cost. Moreover, renewable energy resources need to be explored more for the desorption process to make the DAC process carbon neutral and sustainable. Selecting suitable amine-based adsorbents, which can be regenerated at low temperatures (80–120 °C), may greatly reduce the cost of CO₂ capture.²⁰² The techno-economic analysis of Fasihi et al.³⁵ inferred that the utilization of waste heat for CO₂ desorption can further reduce the CO₂ capture cost by about 40% compared to the lack of free waste heat. The heat required for regeneration of sorbents, can come from a wide range of cost-effective thermal energy supply alternatives, such as heat pumps, geothermal and waste heat generated from some industrial plants such as combined heat and power plants, waste incinerators, electrolysers, or from the synthetic fuel production process.^202,203 The combination of temperature and vacuum swing desorption (TVSD) can be used in practice to produce high purity CO₂,²⁰⁴ which is well commercialized by the Climeworks plant at Hinwil,¹⁹¹ where combined heating and vacuum are used for CO₂ desorption at <100 °C using waste heat developed from municipal waste incineration. Moreover, moisture swing adsorbents such as amine-based ion exchange resin, as reported by Lackner et al., can also be considered as a promising candidate due to its facile low temperature regeneration (45 °C).¹⁷⁸

(v) Further, the selectivity of CO₂ to other bulk components (such as moisture, O₂, and N₂) of amine-based adsorbents needs to be high, so that low parasitic energy²⁰⁵ is required for moisture or N₂ regeneration during the desorption of CO₂ and to produce high purity grade CO₂ (99%) for pipeline transportation (pressure of 10–15 MPa is typically chosen).²⁰⁶

(vi) For moisture-swing adsorption, the used adsorbent, such as ion-exchange resin, can be manufactured with a greater charge density to improve the capacity of moisture-swing sorbents, and materials with varied cation distances can be employed under different humidity conditions.⁹

(vii) Finally, the most critical aspect is the cost of CO₂ capture, which depends upon all the above-discussed milestones. A high sorbent CO₂ uptake capacity and high CO₂/other gas selectivity can reduce the cost of CO₂ capture by reducing the amount of sorbent required.¹⁹⁸ Research should be focused on gathering reliable data on the techno-economic and life-cycle performance of DAC under the actual conditions to predict the environmental and economic viability.¹⁵ Furthermore, incentives for CO₂ storage should be encouraged, such as the 45Q tax credit in the US, which will stimulate learning and cost reductions of DAC at this early stage of its development.²⁰⁷

4 Summary and outlook

CO₂ capture from the air is the need of the hour to combat anthropogenic CO₂ emissions from small and distributed sources, including commercial/domestic fuel-based thermal appliances and the transportation sector. The ability of DAC to harvest CO₂ directly from the atmosphere and sequester it onboard without any need for sophisticated infrastructure and prolonged transportation to recycling sites creates a new era for suitable material and process development. Amine-based materials have shown promising characteristics to be used as suitable adsorbents for DAC, owing to their high CO₂ adsorption capacity at ultra-dilute CO₂ concentration, excellent CO₂ over other components selectivity, and high adsorption under humid conditions, and lower regeneration energy corresponding to liquid amines. Moreover, amine-based adsorbents can be synthesized in a variety of processed including impregnation, grafting, in situ polymerization, and double functionalization techniques. Advantageously, the textural properties and nature of the amine-based adsorbents have been extensively optimized and established to define their suitability for CO₂ captures, such as their giant surface area (for better dispersion amine without clogging), large pore volume (to accommodate more amine moieties), hierarchal pore structure with interconnected pore opening (to facilitate better amine mobility without steric restrictions and faster CO₂ kinetics), enriched surface silanol group (for occluding more amine species by chemical grafting) and so on.

Further, DAC performance depends on the amine loading, amine type, amine density, and its basicity. Impregnating more reactive amine species (primary amine) into the pore improves CO₂ uptake until the percolation is achieved. The preliminary evaluation parameters for optimal sorbent design and their future outlooks have been discussed in this review. Exploration of amine-modified MOF-based adsorbents via in situ polymerization and double functionalization techniques, and their techno-economic analyses needs more attention. Furthermore, to reduce the heat and mass transfer limitation, the regeneration enthalpy of CO₂ of amine-based adsorbents, and the cost of CO₂ capture, innovative regeneration techniques such as moisture swing, electro swing, and magnetic swing need to be further explored and implemented at commercial level. Overall, sorbent design is a crucial factor in DAC, which still needs to be explored further to govern the cost of the DAC process. In the best-case scenario, with extensive and focused efforts, the long-term cost of CO₂ removal using the DAC technique can be reduced from its existing price to 40-80 $ per tCO₂. Aggressive research and innovation in DAC would undoubtedly slow down catastrophic global climate change, resulting in a better living environment.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

Authors thank IIT Indore and SERB New Delhi for the funding support.

References

1. O. Edenhofer, Y. Sokona, J. C. Minx, E. Farahani, S. Kadner, K. Seyboth, A. Adler, I. Baum, S. Brunner, B. Kriemann, J. Savolainen, S. Schlömer, C. von Stechow and T. Zwickel, Climate Change 2014 Mitigation of Climate Change Working Group III Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change Edited by, 2014.
2. O. T. Qazvini and S. G. Telfer, J. Mater. Chem. A, 2020, 8, 12028–12034.
3. S. Chu, Y. Cui and N. Liu, Nat. Mater., 2016, 16, 16–22.
4. Global Monitoring Laboratory - Carbon Cycle Greenhouse Gases, https://gml.noaa.gov/ccgg/trends/, (accessed 28 October 2022).
5. M. Ozkan, S. P. Nayak, A. D. Ruiz and W. Jiang, iScience, 2022, 25, 103990.
6. Net Zero Coalition|United Nations, https://www.un.org/en/climatechange/net-zero-coalition, (accessed 28 October 2022).
7. U.S. Global Change Research Program, U.S. Glob. Chang. Res. Progr., 2018, 1, p. 470.
8. National Academies of Sciences, Engineering, and Medicine, Negative Emissions Technologies and Reliable Sequestration: A Research Agenda, The National Academies Press, Washington, DC, 2019,  DOI:10.17226/25259.
9. X. Shi, H. Xiao, H. Azarabadi, J. Song, X. Wu, X. Chen and K. S. Lackner, Angew. Chem., Int. Ed., 2020, 59, 6984–7006.
10. C. Beuttler, L. Charles and J. Wurzbacher, Front. Clim., 2019, 1, 10.
11. D. Goud, R. Gupta, R. Maligal-Ganesh and S. C. Peter, ACS Catal., 2020, 10, 14258–14282.
12. D. W. Keith, M. Ha-Duong and J. K. Stolaroff, Clim. Change, 2006, 74, 17–45.
13. Can carbon capture facilities reverse climate change?|Climate Crisis News|Al Jazeera, https://www.aljazeera.com/news/2021/11/1/cop, (accessed 11 December 2021).
14. Seizing the Moment – Why Carbon Capture is so Important for COP26 - Green Recruitment Company, https://www.greenrecruitmentcompany.com/blog/2021/11/seizing-the-moment-why-carbon-capture-is-so-important-for-cop26, (accessed 11 December 2021).
15. M. Erans, E. S. Sanz-Pérez, D. P. Hanak, Z. Clulow, D. M. Reiner and G. A. Mutch, Energy Environ. Sci., 2022, 15, 1360–1405.
16. E. S. Sanz-Pérez, C. R. Murdock, S. A. Didas and C. W. Jones, Chem. Rev., 2016, 116, 11840–11876.
17. P. Luis, Desalination, 2016, 380, 93–99.
18. B. R. Strazisar, R. R. Anderson and C. M. White, Energy Fuels, 2003, 17, 1034–1039.
19. A. Chakma and A. Meisen, J. Chromatogr. A, 1988, 457, 287–297.
20. K. Z. House, A. C. Baclig, M. Ranjan, E. A. Van Nierop, J. Wilcox and H. J. Herzog, Proc. Natl. Acad. Sci. U. S. A., 2011, 108, 20428–20433.
21. A. Kiani, K. Jiang and P. Feron, Front. Energy Res., 2020, 8, 92.
22. R. Zhang, T. Li, Y. Zhang, J. Ha, Y. Xiao, C. Li, X. Zhang and H. Luo, Sep. Purif. Technol., 2022, 300, 121702.
23. X. Hu, J. Huang, X. He, Q. Luo, C. Li, C. Zhou and R. Zhang, Fuel, 2022, 316, 123216.
24. G. Chen, G. Chen, M. Peruzzini, R. Zhang and F. Barzagli, Sep. Purif. Technol., 2022, 291, 120939.
25. R. Zhang, X. He, T. Liu, C. Li, M. Xiao, H. Ling, X. Hu, X. Zhang, F. Tang and H. Luo, Sep. Purif. Technol., 2022, 295, 121292.
26. S. Shayegh, V. Bosetti and M. Tavoni, Front. Clim., 2021, 3, 46.
27. F. Zeman, Environ. Sci. Technol., 2007, 41, 7558–7563.
28. A. Bandi, M. Specht, T. Weimer and K. Schaber, Energy Convers. Manage., 1995, 36, 899–902.
29. R. Socolow, M. Desmond, R. Aines, J. Blackstock, O. Bolland, T. Kaarsberg, N. Lewis, M. Mazzotti, A. Pfeffer, K. Sawyer, J. Siirola, B. Smit and J. Wilcox, Direct Air Capture of CO₂ with Chemicals: A Technology Assessment for the APS Panel on Public Affairs, American Physical Society.
30. S. Choi, J. H. Drese and C. W. Jones, ChemSusChem, 2009, 2, 796–854.
31. E. E. Ünveren, B. Ö. Monkul, Ş. Sarıoğlan, N. Karademir and E. Alper, Petroleum, 2017, 3, 37–50.
32. A. Cherevotan, J. Raj and S. C. Peter, J. Mater. Chem. A, 2021, 9, 27271–27303.
33. Y. Belmabkhout, R. Serna-Guerrero and A. Sayari, Ind. Eng. Chem. Res., 2010, 49, 359–365.
34. S. Satyapal, T. Filburn, J. Trela and J. Strange, Energy Fuels, 2001, 15, 250–255.
35. M. Fasihi, O. Efimova and C. Breyer, J. Cleaner Prod., 2019, 224, 957–980.
36. F. Sabatino, A. Grimm, F. Gallucci, M. van Sint Annaland, G. J. Kramer and M. Gazzani, Joule, 2021, 5, 2047–2076.
37. A. Gambhir and M. Tavoni, One Earth, 2019, 1, 405–409.
38. H. Thakkar, A. Issa, A. A. Rownaghi and F. Rezaei, Chem. Eng. Technol., 2017, 40, 1999–2007.
39. H. T. Kwon, M. A. Sakwa-Novak, S. H. Pang, A. R. Sujan, E. W. Ping and C. W. Jones, Chem. Mater., 2019, 31, 5229–5237.
40. A. Goeppert, H. Zhang, R. Sen, H. Dang and G. K. S. Prakash, ChemSusChem, 2019, 12, 1712–1723.
41. D. R. Kumar, C. Rosu, A. R. Sujan, M. A. Sakwa-Novak, E. W. Ping and C. W. Jones, ACS Sustainable Chem. Eng., 2020, 8, 10971–10982.
42. T. M. McDonald, W. R. Lee, J. A. Mason, B. M. Wiers, C. S. Hong and J. R. Long, J. Am. Chem. Soc., 2012, 134, 7056–7065.
43. G. Rim, F. Kong, M. Song, C. Rosu, P. Priyadarshini, R. P. Lively and C. W. Jones, JACS Au, 2022, 2, 380–393.
44. B. Lv, B. Guo, Z. Zhou and G. Jing, Environ. Sci. Technol., 2015, 49, 10728–10735.
45. S. Kim, H. Shi and J. Y. Lee, Int. J. Greenhouse Gas Control, 2016, 45, 181–188.
46. H. Bin Xie, Y. Zhou, Y. Zhang and J. K. Johnson, J. Phys. Chem. A, 2010, 114, 11844–11852.
47. M. Caplow, J. Am. Chem. Soc., 1968, 90, 6795–6803.
48. P. V. Danckwerts, Chem. Eng. Sci., 1979, 34, 443–446.
49. A. Hartono, E. F. da Silva and H. F. Svendsen, Chem. Eng. Sci., 2009, 64, 3205–3213.
50. P. M. M. Blauwhoff, G. F. Versteeg and W. P. M. Van Swaaij, Chem. Eng. Sci., 1984, 39, 207–225.
51. R. B. Said, J. M. Kolle, K. Essalah, B. Tangour and A. Sayari, ACS Omega, 2020, 5, 26125–26133.
52. E. F. Da Silva and H. F. Svendsen, Ind. Eng. Chem. Res., 2004, 43, 3413–3418.
53. P. D. Vaidya and E. Y. Kenig, Chem. Eng. Technol., 2010, 33, 1577–1581.
54. B. Arstad, R. Blom and O. Swang, J. Phys. Chem. A, 2007, 111, 1222–1228.
55. G. F. Versteeg, L. A. J. Van Dijck and W. P. M. Van Swaaij, Chem. Eng. Commun., 1996, 144, 113–158.
56. R. Afonso, M. Sardo, L. Mafra and J. R. B. Gomes, Environ. Sci. Technol., 2019, 53, 2758–2767.
57. P. V. Kortunov, M. Siskin, L. S. Baugh and D. C. Calabro, Energy Fuels, 2015, 29, 5919–5939.
58. S. A. Didas, M. A. Sakwa-Novak, G. S. Foo, C. Sievers and C. W. Jones, J. Phys. Chem. Lett., 2014, 5, 4194–4200.
59. P. V. Kortunov, L. S. Baugh, M. Siskin and D. C. Calabro, Energy Fuels, 2015, 29, 5967–5989.
60. P. V. Kortunov, M. Siskin, M. Paccagnini and H. Thomann, Energy Fuels, 2016, 30, 1223–1236.
61. P. V. Kortunov, M. Siskin, L. S. Baugh and D. C. Calabro, Energy Fuels, 2015, 29, 5940–5966.
62. Y. G. Ko, S. S. Shin and U. S. Choi, J. Colloid Interface Sci., 2011, 361, 594–602.
63. X. Shen, H. Du, R. H. Mullins and R. R. Kommalapati, Energy Technol., 2017, 5, 822–833.
64. T. Čendak, L. Sequeira, M. Sardo, A. Valente, M. L. Pinto and L. Mafra, Chem. – Eur. J., 2018, 24, 10136–10145.
65. M. W. Hahn, M. Steib, A. Jentys and J. A. Lercher, J. Phys. Chem. C, 2015, 119, 4126–4135.
66. M. W. Hahn, J. Jelic, E. Berger, K. Reuter, A. Jentys and J. A. Lercher, J. Phys. Chem. B, 2016, 120, 1988–1995.
67. L. Mafra, T. Čendak, S. Schneider, P. V. Wiper, J. Pires, J. R. B. Gomes and M. L. Pinto, J. Am. Chem. Soc., 2017, 139, 389–408.
68. Y. Zhai and S. S. C. Chuang, Energy Technol., 2017, 5, 510–519.
69. S. A. Didas, R. Zhu, N. A. Brunelli, D. S. Sholl and C. W. Jones, J. Phys. Chem. C, 2014, 118, 12302–12311.
70. A. Sayari, A. Heydari-Gorji and Y. Yang, J. Am. Chem. Soc., 2012, 134, 13834–13842.
71. M. L. Pinto, L. Mafra, J. M. Guil, J. Pires and J. Rocha, Chem. Mater., 2011, 23, 1387–1395.
72. A. Sayari and Y. Belmabkhout, J. Am. Chem. Soc., 2010, 132, 6312–6314.
73. C. H. Chen, D. Shimon, J. J. Lee, F. Mentink-Vigier, I. Hung, C. Sievers, C. W. Jones and S. E. Hayes, J. Am. Chem. Soc., 2018, 140, 8648–8651.
74. R. W. Flaig, T. M. Osborn Popp, A. M. Fracaroli, E. A. Kapustin, M. J. Kalmutzki, R. M. Altamimi, F. Fathieh, J. A. Reimer and O. M. Yaghi, J. Am. Chem. Soc., 2017, 139, 12125–12128.
75. E. G. Moschetta, M. A. Sakwa-Novak, J. L. Greenfield and C. W. Jones, Langmuir, 2015, 31, 2218–2227.
76. F. Inagaki, C. Matsumoto, T. Iwata and C. Mukai, J. Am. Chem. Soc., 2017, 139, 4639–4642.
77. M. Yang, C. Ma, M. Xu, S. Wang and L. Xu, Curr. Pollut. Rep., 2019, 5, 272–293.
78. Y. Deng, J. Li, Y. Miao and D. Izikowitz, Energy Rep., 2021, 7, 3506–3516.
79. N. McQueen, K. V. Gomes, C. McCormick, K. Blumanthal, M. Pisciotta and J. Wilcox, Prog. Energy, 2021, 3, 032001.
80. F. Kong, G. Rim, M. G. Song, C. Rosu, P. Priyadarshini, R. P. Lively, M. J. Realff and C. W. Jones, Korean J. Chem. Eng., 2022, 39, 1–19.
81. A. Goeppert, H. Zhang, M. Czaun, R. B. May, G. K. S. Prakash, G. A. Olah and S. R. Narayanan, ChemSusChem, 2014, 7, 1386–1397.
82. A. Kumar, D. G. Madden, M. Lusi, K. J. Chen, E. A. Daniels, T. Curtin, J. J. Perry and M. J. Zaworotko, Angew. Chem., Int. Ed., 2015, 54, 14372–14377.
83. C. Chen, S. Zhang, K. H. Row and W. S. Ahn, J. Energy Chem., 2017, 26, 868–880.
84. S. A. Didas, A. R. Kulkarni, D. S. Sholl and C. W. Jones, ChemSusChem, 2012, 5, 2058–2064.
85. A. Goeppert, H. Zhang, R. Sen, H. Dang and G. K. S. Prakash, ChemSusChem, 2019, 12, 1712–1723.
86. Y. Miao, Z. He, X. Zhu, D. Izikowitz and J. Li, Chem. Eng. J., 2021, 426, 131875.
87. D. W. F. Brilman and R. Veneman, Energy Procedia, 2013, 37, 6070–6078.
88. S. Choi, M. L. Gray and C. W. Jones, ChemSusChem, 2011, 4, 628–635.
89. W. Chaikittisilp, R. Khunsupat, T. T. Chen and C. W. Jones, Ind. Eng. Chem. Res., 2011, 50, 14203–14210.
90. A. Sayari, Q. Liu and P. Mishra, ChemSusChem, 2016, 9, 2796–2803.
91. M. E. Potter, K. M. Cho, J. J. Lee and C. W. Jones, ChemSusChem, 2017, 10, 2192–2201.
92. W. Chaikittisilp, J. D. Lunn, D. F. Shantz and C. W. Jones, Chem. – Eur. J., 2011, 17, 10556–10561.
93. J. C. Hicks, J. H. Drese, D. J. Fauth, M. L. Gray, G. Qi and C. W. Jones, J. Am. Chem. Soc., 2008, 130, 2902–2903.
94. J. H. Drese, S. Choi, R. P. Lively, W. J. Koros, D. J. Fauth, M. L. Gray and C. W. Jones, Adv. Funct. Mater., 2009, 19, 3821–3832.
95. S. Choi, J. H. Drese, P. M. Eisenberger and C. W. Jones, Environ. Sci. Technol., 2011, 45, 2420–2427.
96. E. J. Acosta, C. S. Carr, E. E. Simanek and D. F. Shantz, Adv. Mater., 2004, 16, 985–989.
97. G. Qi, L. Fu, X. Duan, B. H. Choi, M. Abraham and E. P. Giannelis, Greenhouse Gases: Sci. Technol., 2011, 1, 278–284.
98. N. Shanmugam, K. T. Lee, W. Y. Cheng and S. Y. Lu, J. Polym. Sci., Part A: Polym. Chem., 2012, 50, 2521–2526.
99. X. Wang, L. Chen and Q. Guo, Chem. Eng. J., 2015, 260, 573–581.
100. S. A. Didas, S. Choi, W. Chaikittisilp and C. W. Jones, Acc. Chem. Res., 2015, 48, 2680–2687.
101. A. R. Sujan, S. H. Pang, G. Zhu, C. W. Jones and R. P. Lively, ACS Sustainable Chem. Eng., 2019, 7, 5264–5273.
102. A. Heydari-Gorji and A. Sayari, Ind. Eng. Chem. Res., 2012, 51, 6887–6894.
103. A. Wagner, B. Steen, G. Johansson, E. Zanghellini, P. Jacobsson and P. Johansson, Int. J. Spectrosc., 2013, 2013, 1–8.
104. W. J. Son, J. S. Choi and W. S. Ahn, Microporous Mesoporous Mater., 2008, 113, 31–40.
105. Z. Chen, S. Deng, H. Wei, B. Wang, J. Huang and G. Yu, ACS Appl. Mater. Interfaces, 2013, 5, 6937–6945.
106. J. Wang, H. Huang, M. Wang, L. Yao, W. Qiao, D. Long and L. Ling, Ind. Eng. Chem. Res., 2015, 54, 5319–5327.
107. H. Sehaqui, M. E. Gálvez, V. Becatinni, Y. Cheng Ng, A. Steinfeld, T. Zimmermann and P. Tingaut, Environ. Sci. Technol., 2015, 49, 3167–3174.
108. S. H. Pang, R. P. Lively and C. W. Jones, ChemSusChem, 2018, 11, 2628–2637.
109. A. R. Sujan, D. R. Kumar, M. Sakwa-Novak, E. W. Ping, B. Hu, S. J. Park and C. W. Jones, ACS Appl. Polym. Mater., 2019, 1, 3137–3147.
110. G. Rim, T. G. Feric, T. Moore and A. H. A. Park, Adv. Funct. Mater., 2021, 31, 2010047.
111. K. A. S. Abhilash, T. Deepthi, R. A. Sadhana and K. G. Benny, ACS Appl. Mater. Interfaces, 2015, 7, 17969–17976.
112. J. A. Wurzbacher, C. Gebald and A. Steinfeld, Energy Environ. Sci., 2011, 4, 3584–3592.
113. C. Gebald, J. A. Wurzbacher, P. Tingaut, T. Zimmermann and A. Steinfeld, Environ. Sci. Technol., 2011, 45, 9101–9108.
114. Y. Kong, G. Jiang, Y. Wu, S. Cui and X. Shen, Chem. Eng. J., 2016, 306, 362–368.
115. W. R. Lee, S. Y. Hwang, D. W. Ryu, K. S. Lim, S. S. Han, D. Moon, J. Choi and C. S. Hong, Energy Environ. Sci., 2014, 7, 744–751.
116. T. M. McDonald, J. A. Mason, X. Kong, E. D. Bloch, D. Gygi, A. Dani, V. Crocellà, F. Giordanino, S. O. Odoh, W. S. Drisdell, B. Vlaisavljevich, A. L. Dzubak, R. Poloni, S. K. Schnell, N. Planas, K. Lee, T. Pascal, L. F. Wan, D. Prendergast, J. B. Neaton, B. Smit, J. B. Kortright, L. Gagliardi, S. Bordiga, J. A. Reimer and J. R. Long, Nature, 2015, 519, 303–308.
117. N. Planas, A. L. Dzubak, R. Poloni, L. C. Lin, A. McManus, T. M. McDonald, J. B. Neaton, J. R. Long, B. Smit and L. Gagliardi, J. Am. Chem. Soc., 2013, 135, 7402–7405.
118. P. Q. Liao, X. W. Chen, S. Y. Liu, X. Y. Li, Y. T. Xu, M. Tang, Z. Rui, H. Ji, J. P. Zhang and X. M. Chen, Chem. Sci., 2016, 7, 6528–6533.
119. C. Gebald, J. A. Wurzbacher, A. Borgschulte, T. Zimmermann and A. Steinfeld, Environ. Sci. Technol., 2014, 48, 2497–2504.
120. W. Lu, J. P. Sculley, D. Yuan, R. Krishna and H. C. Zhou, J. Phys. Chem. C, 2013, 117, 4057–4061.
121. G. Qi, L. Fu and E. P. Giannelis, Nat. Commun., 2014, 5, 1–7.
122. H. Lyu, H. Li, N. Hanikel, K. Wang and O. M. Yaghi, J. Am. Chem. Soc., 2022, 2022, 12989–12995.
123. A. Goeppert, M. Czaun, R. B. May, G. K. S. Prakash, G. A. Olah and S. R. Narayanan, J. Am. Chem. Soc., 2011, 133, 20164–20167.
124. W. Zhang, H. Liu, C. Sun, T. C. Drage and C. E. Snape, Chem. Eng. Sci., 2014, 116, 306–316.
125. W. Chaikittisilp, H. J. Kim and C. W. Jones, Energy Fuels, 2011, 25, 5528–5537.
126. M. A. Sakwa-Novak and C. W. Jones, ACS Appl. Mater. Interfaces, 2014, 6, 9245–9255.
127. X. Wang, X. Ma, V. Schwartz, J. C. Clark, S. H. Overbury, S. Zhao, X. Xu and C. Song, Phys. Chem. Chem. Phys., 2012, 14, 1485–1492.
128. M. A. Sakwa-Novak, S. Tan and C. W. Jones, ACS Appl. Mater. Interfaces, 2015, 7, 24748–24759.
129. M. L. Sarazen, M. A. Sakwa-Novak, E. W. Ping and C. W. Jones, ACS Sustainable Chem. Eng., 2019, 7, 7338–7345.
130. A. Holewinski, M. A. Sakwa-Novak and C. W. Jones, J. Am. Chem. Soc., 2015, 137, 11749–11759.
131. C. Rosu, S. H. Pang, A. R. Sujan, M. A. Sakwa-Novak, E. W. Ping and C. W. Jones, ACS Appl. Mater. Interfaces, 2020, 12, 38085–38097.
132. S. H. Pang, L. C. Lee, M. A. Sakwa-Novak, R. P. Lively and C. W. Jones, J. Am. Chem. Soc., 2017, 139, 3627–3630.
133. Y. Kuwahara, D. Y. Kang, J. R. Copeland, N. A. Brunelli, S. A. Didas, P. Bollini, C. Sievers, T. Kamegawa, H. Yamashita and C. W. Jones, J. Am. Chem. Soc., 2012, 134, 10757–10760.
134. Y. Kuwahara, D.-Y. Kang, J. R. Copeland, P. Bollini, C. Sievers, T. Kamegawa, H. Yamashita and C. W. Jones, Chem. – Eur. J., 2012, 18, 16649–16664.
135. L. A. Darunte, A. D. Oetomo, K. S. Walton, D. S. Sholl and C. W. Jones, ACS Sustainable Chem. Eng., 2016, 4, 5761–5768.
136. X. Xu, M. B. Myers, F. G. Versteeg, B. Pejcic, C. Heath and C. D. Wood, Chem. Commun., 2020, 56, 7151–7154.
137. L. He, M. Fan, B. Dutcher, S. Cui, X. Dong Shen, Y. Kong, A. G. Russell and P. McCurdy, Chem. Eng. J., 2012, 189–190, 13–23.
138. N. R. Stuckert and R. T. Yang, Environ. Sci. Technol., 2011, 45, 10257–10264.
139. S. Choi, T. Watanabe, T. H. Bae, D. S. Sholl and C. W. Jones, J. Phys. Chem. Lett., 2012, 3, 1136–1141.
140. H. Li, K. Wang, D. Feng, Y. P. Chen, W. Verdegaal and H. C. Zhou, ChemSusChem, 2016, 9, 2832–2840.
141. F. Q. Liu, L. Wang, Z. G. Huang, C. Q. Li, W. Li, R. X. Li and W. H. Li, ACS Appl. Mater. Interfaces, 2014, 6, 4371–4381.
142. V. Nikulshina, C. Gebald and A. Steinfeld, Chem. Eng. J., 2009, 146, 244–248.
143. Y. Belmabkhout, R. Serna-Guerrero and A. Sayari, Chem. Eng. Sci., 2010, 65, 3695–3698.
144. A. K. Voice and G. T. Rochelle, Ind. Eng. Chem. Res., 2014, 53, 16222–16228.
145. T. Wang and K. J. Jens, Energy Procedia, 2012, 23, 102–110.
146. S. Chi and G. T. Rochelle, Ind. Eng. Chem. Res., 2002, 41, 4178–4186.
147. M. Jahandar Lashaki, S. Khiavi and A. Sayari, Chem. Soc. Rev., 2019, 48, 3320–3405.
148. A. Ahmadalinezhad and A. Sayari, Phys. Chem. Chem. Phys., 2013, 16, 1529–1535.
149. A. Heydari-Gorji, Y. Belmabkhout and A. Sayari, Microporous Mesoporous Mater., 2011, 145, 146–149.
150. H. Zhang, A. Goeppert, G. K. S. Prakash and G. Olah, RSC Adv., 2015, 5, 52550–52562.
151. S. Bali, T. T. Chen, W. Chaikittisilp and C. W. Jones, Energy Fuels, 2013, 27, 1547–1554.
152. C. S. Srikanth and S. S. C. Chuang, ChemSusChem, 2012, 5, 1435–1442.
153. C. S. Srikanth and S. S. C. Chuang, J. Phys. Chem. C, 2013, 117, 9196–9205.
154. W. Choi, K. Min, C. Kim, Y. S. Ko, J. W. Jeon, H. Seo, Y.-K. Park and M. Choi, Nat. Commun., 2016, 7, 12640.
155. A. R. Kulkarni and D. S. Sholl, Ind. Eng. Chem. Res., 2012, 51, 8631–8645.
156. L. A. Darunte, K. S. Walton, D. S. Sholl and C. W. Jones, Curr. Opin. Chem. Eng., 2016, 12, 82–90.
157. S. Keskin, T. M. van Heest and D. S. Sholl, ChemSusChem, 2010, 3, 879–891.
158. P. Saha and S. Chowdhury, Insights Into Thermodynamics, ed. M. Tadashi, INTECH Open Acess Publisher, 2011, ch. 16, pp. 349–364.
159. A. L. Myers, AIChE J., 2002, 48, 145–160.
160. J. C. Moreno-Piraján and L. Giraldo, J. Chem. Eng. Data, 2020, 65, 3130–3145.
161. A. Nuhnen and C. Janiak, Dalton Trans., 2020, 49, 10295–10307.
162. D. Panda, E. A. Kumar and S. K. Singh, J. CO2 Util., 2020, 40, 101223.
163. M. E. Potter, S. H. Pang and C. W. Jones, Langmuir, 2016, 33, 117–124.
164. M. A. Alkhabbaz, P. Bollini, G. S. Foo, C. Sievers and C. W. Jones, J. Am. Chem. Soc., 2014, 136, 13170–13173.
165. C.-J. Yoo, L.-C. Lee and C. W. Jones, Langmuir, 2015, 31, 13350–13360.
166. A. L. Kohl and R. B. Nielsen, Alkanolamines for Hydrogen Sulfide and Carbon Dioxide Removal, ed. A. L. Kohl and R. B. Nielsen, Gulf Professional Publishing, Houston, 1997, ch. 2, pp. 40–186.
167. A. Goeppert, S. Meth, G. K. S. Prakash and G. A. Olah, Energy Environ. Sci., 2010, 3, 1949–1960.
168. J. A. Wurzbacher, C. Gebald, N. Piatkowski and A. Steinfeld, Environ. Sci. Technol., 2012, 46, 9191–9198.
169. A. A. Al-Absi, M. Mohamedali, A. Domin, A. M. Benneker and N. Mahinpey, Chem. Eng. J., 2022, 447, 137465.
170. B. Yeskendir, J. P. Dacquin, Y. Lorgouilloux, C. Courtois, S. Royer and J. Dhainaut, Mater. Adv., 2021, 2, 7139–7186.
171. F. Rezaei and P. Webley, Sep. Purif. Technol., 2010, 70, 243–256.
172. S. Lawson, X. Li, H. Thakkar, A. A. Rownaghi and F. Rezaei, Chem. Rev., 2021, 121, 6246–6291.
173. H. Yang, X. Fang, Z. Li, H. Sun and H. Chen, ACS Sustainable Chem. Eng., 2022, 10, 1759–1764.
174. R. M. Heck, S. Gulati and R. J. Farrauto, Chem. Eng. J., 2001, 82, 149–156.
175. L. A. Darunte, Y. Terada, C. R. Murdock, K. S. Walton, D. S. Sholl and C. W. Jones, ACS Appl. Mater. Interfaces, 2017, 9, 17042–17050.
176. M. A. Sakwa-Novak, C. J. Yoo, S. Tan, F. Rashidi and C. W. Jones, ChemSusChem, 2016, 9, 1859–1868.
177. R. P. Wijesiri, G. P. Knowles, H. Yeasmin, A. F. A. Hoadley and A. L. Chaffee, Ind. Eng. Chem. Res., 2019, 58, 3293–3303.
178. K. S. Lackner, Eur. Phys. J.: Spec. Top., 2009, 176, 93–106.
179. H. He, W. Li, M. Zhong, D. Konkolewicz, D. Wu, K. Yaccato, T. Rappold, G. Sugar, N. E. David and K. Matyjaszewski, Energy Environ. Sci., 2013, 6, 488–493.
180. T. Wang, J. Liu, M. Fang and Z. Luo, Energy Procedia, 2013, 37, 6096–6104.
181. X. Shi, H. Xiao, K. S. Lackner and X. Chen, Angew. Chem., Int. Ed., 2016, 55, 4026–4029.
182. S. Rost, D. Gerten, A. Bondeau, W. Lucht, J. Rohwer and S. Schaphoff, Water Resour. Res., 2008, 44, 9405.
183. T. Wang, K. S. Lackner and A. B. Wright, Phys. Chem. Chem. Phys., 2013, 15, 504–514.
184. H. He, W. Li, M. Lamson, M. Zhong, D. Konkolewicz, C. M. Hui, K. Yaccato, T. Rappold, G. Sugar, N. E. David, K. Damodaran, S. Natesakhawat, H. Nulwala and K. Matyjaszewski, Polymer, 2014, 55, 385–394.
185. H. He, M. Zhong, D. Konkolewicz, K. Yacatto, T. Rappold, G. Sugar, N. E. David and K. Matyjaszewski, J. Mater. Chem. A, 2013, 1, 6810–6821.
186. H. He, M. Zhong, D. Konkolewicz, K. Yacatto, T. Rappold, G. Sugar, N. E. David, J. Gelb, N. Kotwal, A. Merkle and K. Matyjaszewski, Adv. Funct. Mater., 2013, 23, 4720–4728.
187. T. Wang, K. S. Lackner and A. Wright, Environ. Sci. Technol., 2011, 45, 6670–6675.
188. C. Hou, Y. Wu, T. Wang, X. Wang and X. Gao, Energy Fuels, 2018, 33, 1745–1752.
189. J. Song, J. Liu, W. Zhao, Y. Chen, H. Xiao, X. Shi, Y. Liu and X. Chen, Ind. Eng. Chem. Res., 2018, 57, 4941–4948.
190. G. Holmes and D. W. Keith, Philos. Trans. R. Soc., A, 2012, 370, 4380–4403.
191. S. Deutz and A. Bardow, Nat. Energy, 2021, 6, 203–213.
192. Direct air capture technology and carbon removal, https://climeworks.com/co2-removal, (accessed 7 May 2022).
193. E. Ping, M. Sakwa-Novak and P. Eisenberger, in International Conference on Negative CO₂ Emissions, Gothenburg, 2018.
194. A. Sinha, L. A. Darunte, C. W. Jones, M. J. Realff and Y. Kawajiri, Ind. Eng. Chem. Res., 2017, 56, 750–764.
195. H. Azarabadi and K. S. Lackner, Appl. Energy, 2019, 250, 959–975.
196. M. M. Sadiq, M. P. Batten, X. Mulet, C. Freeman, K. Konstas, J. I. Mardel, J. Tanner, D. Ng, X. Wang, S. Howard, M. R. Hill and A. W. Thornton, Adv. Sustainable Syst., 2020, 4, 2000101.
197. C. J. E. Bajamundi, J. Koponen, V. Ruuskanen, J. Elfving, A. Kosonen, J. Kauppinen and J. Ahola, J. CO2 Util., 2019, 30, 232–239.
198. H. A. Patel, J. Byun and C. T. Yavuz, ChemSusChem, 2017, 10, 1303–1317.
199. J. Park, H. O. Rubiera Landa, Y. Kawajiri, M. J. Realff, R. P. Lively and D. S. Sholl, Ind. Eng. Chem. Res., 2020, 59, 7097–7108.
200. Y. S. Bae and R. Q. Snurr, Angew. Chem., Int. Ed., 2011, 50, 11586–11596.
201. E. S. M. El-Sayed and D. Yuan, Green Chem., 2020, 22, 4082–4104.
202. Y. Qiu, P. Lamers, V. Daioglou, N. McQueen, H. S. de Boer, M. Harmsen, J. Wilcox, A. Bardow and S. Suh, Nat. Commun., 2022, 13, 1–13.
203. C. Breyer, M. Fasihi, C. Bajamundi and F. Creutzig, Joule, 2019, 3, 2053–2057.
204. J. Elfving, C. Bajamundi, J. Kauppinen and T. Sainio, J. CO2 Util., 2017, 22, 270–277.
205. M. Oschatz and M. Antonietti, Energy Environ. Sci., 2018, 11, 57–70.
206. M. Anheden, A. Andersson, C. Bernstone, S. Eriksson, J. Yan, S. Liljemark and C. Wall, Greenh. Gas Control Technol., 2005, pp. 2559–2564.
207. G. F. Nemet, M. W. Callaghan, F. Creutzig, S. Fuss, J. Hartmann, J. Hilaire, W. F. Lamb, J. C. Minx, S. Rogers and P. Smith, Environ. Res. Lett., 2018, 13, 063003.

---

Footnote

† Authors with equal contribution.

---