Improving CO₂ adsorption efficiency of an amine-modified MOF-808 through the synthesis of its graphene oxide composites

Abstract

This research developed a novel composite of MOF-NH₂ and graphene oxide (GO) for enhanced CO₂ capture. Employing the response surface methodology-central composite design (RSM-CCD) for experiments design, various MOF-NH₂/GO samples with GO loadings from 0 to 30 wt% were synthesized. The results of SEM, XRD, EDS, and BET analysis revealed that the materials maintained their MOF crystal structure, confirmed by X-ray diffraction, and exhibited unique texture, high porosity, and oxygen-enriched surface chemistry. The influence of temperature (25–65 °C) and pressure (1–9 bar) on CO₂ adsorption capacity was assessed using a volumetric adsorption system. Optimum conditions were obtained at weight percent of 22.6 wt% GO, temperature of 25 °C, and pressure of 9 bar with maximum adsorption capacity of 303.61 mg/g. The incorporation of amino groups enhanced the CO₂ adsorption capacity. Isotherm and kinetic analyses indicated that Freundlich and Fractional-order models best described CO₂ adsorption behavior. Thermodynamic analysis showed the process was exothermic, spontaneous, and physical, with enthalpy changes of − 16.905 kJ/mol, entropy changes of − 0.030 kJ/mol K, and Gibs changes energy of − 7.904 kJ/mol. Mass transfer diffusion coefficients increased with higher GO loadings. Regenerability tests demonstrated high performance and resilience, with only a 5.79% decrease in efficiency after fifteen cycles. These findings suggest significant potential for these composites in CO₂ capture technologies.

Introduction

The combustion of fossil fuels leads to an overabundance of carbon dioxide (CO₂) emissions, which in turn gives rise to significant global climate and environmental issues. Consequently, a growing focus has been on carbon capture and storage (CCS) technology worldwide¹. Using liquid amines for CO₂ capture and storage has been extensively researched and expanded over several decades. This approach has proven viable and effective technology, leading to widespread implementation across several industrial sectors². Nevertheless, this approach presents several limitations, such as the requirement for high energy usage for recovery, potential corrosion-related issues, and the loss of liquid amines during absorption³. Recent advances include the development of amine-based nanofluids and non-aqueous amine solutions, which improve absorption efficiency and minimize energy consumption. The inclusion of nanoparticles in amine-based nanofluids results in increased absorption capacities and enhanced mass transfer performance, making them a potential alternative to standard solvents⁴. Water-lean amines, such as alkoxy-functionalized amines, provide lower viscosity and more energy-efficient regeneration while retaining high CO₂ capture efficiency⁵. However, these novel methods are not without drawbacks, such as high costs and potential stability problems⁶.

In order to tackle the concerns mentioned earlier, scientists have paid significant attention to the advancement of robust sorbents to capture CO₂ due to their advantageous characteristics, such as reduced energy demands for recovery, enhanced capacity for adsorption, selective behaviors in gas uptake, and stable efficiency across adsorption–desorption cycles⁷. Typically, the adsorption techniques employ solid sorbents, including porous organic polymers (POPs)⁸, zeolites⁹, activated carbon (AC)¹⁰, and metal organic frameworks (MOFs)^11,12. Metal–organic frameworks (MOFs) are a kind of porous, crystalline material that are utilized in a wide range of applications. They can be used for a variety of purposes, such as gas adsorption, heterogeneous catalysis, ion exchange, and molecular separation, with reduced adsorbent volume and mass than traditional adsorbents because of their high tunability in design, ultrahigh surface areas, and wide structural and chemical diversity^13,14. The most appealing feature of these materials is the presence of open metal sites or coordinatively unsaturated metal nodes along the pore surfaces. These cations of five-coordinate metals may be tuned to undergo post-synthetic functionalization, acting as Lewis acids and significantly polarizing gas adsorbents. Because of their extremely high porosity, regular porous structures, and chemical functionalities that may be altered by varying the organic linker or metallic center, MOFs have been proposed as potential candidates for CO₂ capture¹⁵. Wenger et al. demonstrated the use of redox-active MOFs for electrochemical CO₂ capture, offering a less energy-intensive alternative to traditional methods¹⁶. Gebremariam et al. highlighted the hybridization of MOFs with other materials to improve stability and performance, addressing issues such as poor thermal stability and sensitivity to contaminants¹⁷. Qiu et al. presented the combination of MOFs with superbase-derived ionic liquids for efficient direct air capture, showcasing enhanced CO₂ capture kinetics and stability¹⁸. The reviewed research illustrate how modifying the MOFs with amines enhances their ability to adsorb CO₂. Dinda reported that rafting of diethylenetriamine (DETA) into zirconium fumarate MOFs resulted in a maximum CO₂ adsorption capacity of 86 mg/g for DETA-modified ZrFu-MOF at 30 °C, compared to 73 mg/g for the pristine ZrFu-MOF¹⁹. Jo et al. presented a Cu-based ultramicroporous MOF, with a CO₂ adsorption capacity of 88.44 mg/g at 298 K and 15 kPa, which was retained its performance under humid conditions²⁰. Jun et al. highlighted TEPA-functionalized MOF-808, which exhibited a CO₂ adsorption capacity of approximately 2.5 times that of the pristine MOF-808 at 15 kPa²¹. Gaikwad et al. developed an amine-modified MOF-177 with a significant improvement in CO₂ adsorption capacity, obtained about 202.4 mg/g at 328 K, which was substantially higher than the parent MOF-177²².

Despite the intriguing characteristics, several MOFs demonstrate vulnerabilities preventing widespread implementation. For example, low water stability, while it has improved recently, is still a big problem and leads to the collapse of materials that have lost their crystallinity²³. The fragility of the metal/ligand connections is frequently cited as the underlying cause for this phenomenon. Furthermore, because to the presence of an organic component, MOFs exhibit lower thermal stability compared to zeolites, which are their structural counterparts. The average temperature range at which MOFs undergo breakdown is between 300 and 500 °C. In contrast to many other porous materials, such activated carbons, MOFs do not have atomically dense pores. In general, MOFs exhibit either predominantly microporous or mesoporous characteristics. However, for the purpose of small molecule adsorption, a hierarchical arrangement of mesopores and micropores is frequently preferred. In fact, mesopores improve kinetics and assist diffusion for separation purposes, while micropores allow for more adsorption capacity^24,25.

Typically two methods are commonly suggested to tackle the issues mentioned earlier. The first technique involves modifying the MOF by functionalizing the ligands and adjusting the metal and ligand interactions. Since their structures and features are extensively adjustable, moisture-resistant MOFs can be created by carefully choosing building blocks, for example, by utilizing ligands with a high pKa value²⁶. The second strategy entails constructing MOFs based composite samples such as MOF/alumina²⁷, MOF/polymer nanoparticles²⁸, MOF/carbon nanotube²⁹, and MOF/graphene oxides (GO)^24,30. The latter strategy has significant appeal due to its ability to mitigate the limitations of the MOF and yield synergistic features³¹. Graphene oxide (GO) is a carbon-based substance that possesses a two-dimensional structure and is abundant in functional groups, including carboxyl, hydroxyl, and epoxy functional groups. It exhibits notable attributes such as a substantial specific surface area, pronounced hydrophilicity, and commendable chemical stability. Thus, MOF composites are promising due to their potential benefits³². The use of GO/MOF composites has significantly enhanced the adaptability of MOFs, leading to the emergence of novel applications across diverse disciplines³³. The GO has a high atom density, possesses several functional groups, and can induce the crystallization of MOFs under ambient conditions. Hence, the characterization of the interaction between the GO and the MOF in a GO/MOF composite material is heavily influenced by the specific conditions and driving forces utilized throughout the synthesis process^34,35. Furthermore, it is worth noting that the morphology and dimensions of the MOF play a crucial role in determining the effectiveness of the composites in various targeted applications³⁶. Size management is a crucial approach to modulating the mechanical characteristics of crystalline materials. The presence of oxidized groups on GO’s surfaces and oxygen atoms on the MOF’s ligand might lead to a competitive binding process with the metallic ions, so that this competition may result in the formation of free MOF particles. There are some conventional synthesis methods for the preparation of the MOF/GO composites, such as in-situ growth³⁷, self-assembly method³⁸, Pickering emulsion³¹, and mechanical mixing³⁹, among them, in-situ growth technique is attracted more attention due to its simple, direct, and rapid synthesis procedure²⁵.

In recent years, some studies on gas adsorption using MOF/GO materials resulted in an enhancement in the composite’s gas uptake capabilities compared to its parent MOF. Bandose et al.⁴⁰ prepared a MOF/GO composite sample using Cu-based MOF (HKUST-1) and investigated its CO₂ uptake capability. The results shown that the CO₂ uptake capacity of the composite was 186.12 mg·g⁻¹ at the atmospheric pressure and 30 °C, which was twice more than the parent MOF. Cao et al.⁴¹ synthesized and studied the effect of GO’s loading on the UiO-66/GO composite. The results showed that the optimum GO loading quantity was 5 wt%, with the highest CO₂ adsorption capacity of 154.44 mg·g⁻¹ under ambient conditions. In another research conducted by Cao et al. a novel aminated UiO-66-NH₂/GO composite was developed and the CO₂ uptake capacity of 282.04 mg·g⁻¹ was obtained for composite sample which was higher than for pristine composite UiO-66/GO (268.4 mg·g⁻¹)⁴². Lin et al. conducted a computer modeling study on the MOF-5/GO material, whereby they performed calculations to determine the CO₂ and alkynes adsorption characteristics⁴³. The study demonstrated that the MOF-5/GO interface region contains the binding site responsible for the adsorption enhancement effect. Also, as shown by the study of linear alkanes, the removal of long-chain alkanes from natural gas and the separation of hydrocarbons are possible because of the MOF-5/GO composite material’s superior adsorption enhancing effect on big guest molecules. Ram Kumar et al. developed a ZIF-8/GO nanocomposite with adjustable porosity and high CO₂ uptake capacity. The GO’s sheet serves as a substrate for the formation and enlargement of nanocrystals⁴⁴. Additionally, it works as a guiding agent for the controlled development of ZIF-8 nanocrystals by modulating their coordination. The nano-scale shape of the composite progressively shifts from hexagonal to spherical as the concentration of GO increases. Zarei Mohammadabad et al. prepared a HKUST-1@GrO composite with the maximum CO₂ adsorption capacity of 547.36 mg·g⁻¹ at 283 K and 10 bar that was attributed to increased surface area and specific interactions between the MOF and graphene oxide⁴⁵. Ren et al. developed a novel Cu-BTC/GA-IL composite which exhibited a higher CO₂ uptake capability compared to Cu-BTC/GA without IL additives⁴⁶. Ning et al. synthesized a PI-UiO/GO-1 composite that demonstrated a CO₂ capacity of 362.56 mg·g⁻¹ at 298 K and 30 bar⁴⁷.

Among the several MOF samples investigated for diverse purposes in engineering, the MOF-808 received significant interest due to its multiple features, especially regarding its potential for CO₂ capture. Firstly, it is noteworthy that MOF-808 has a notable capacity for gas adsorption, specifically in the case of CO₂ uptake. These materials’ superior CO₂ uptake capabilities may be attributed to their substantial surface area and pore volume. In addition, in comparison with other MOFs, MOF-808 exhibits exceptional stability and compatibility, guaranteeing the preservation of its structural integrity and adsorption capabilities. Furthermore, the facile synthetic techniques employed in the production of MOF-808 have enabled the incorporation of amine functional moieties, hence enabling convenient modifications. The use of amine-based ligands and the GO in the synthesis process improves the specificity of CO₂ adsorption due to their high affinity for CO₂ molecules through both physisorption and chemisorption mechanisms. Finally, the study fills a knowledge gap in developing an amine-modified MOF/GO composite by investigating the interaction between the composite and the CO₂ molecules, especially at varying weight percentages of the GO⁴⁸.

This study aims to provide innovative adsorbents for CO₂ capture in various applications. To do this, an in-situ synthesis technique was applied to prepare an amine-modified MOF-808/GO composite. This manipulation alters the parent MOF’s structural characteristics and surface chemistry. In order to do this, the Zr-BTC MOF with an optimum weight percentage value of an NH₂ ligand obtained previously by Esfahani et al. was modified by introducing different loading values of the GO⁴⁹. As a result, a unique MOF-NH₂/GO sample was synthesized, and the sample exhibited the maximum capacity for CO₂ adsorption. Furthermore, the elemental composition and morphological features of the resultant samples are examined by the use of several analytical techniques, including Fourier Transform Infrared Spectroscopy (FTIR), Scanning Electron Microscopy (SEM), Energy-Dispersive X-ray Spectroscopy (EDS), and Brunauer–Emmett–Teller (BET) analysis. In addition, due to the lack of a predictive model for the CO₂ uptake capability of the resultant composite, a semi-empirical model is provided based on the RSM-CCD approach and also the CO₂ adsorption isotherm and kinetic modeling are conducted. These predictive models are crucial in the design of CO₂ capture plants. Furthermore, the feasibility of the adsorption process is examined by the determination of thermodynamic parameters. Additionally, the diffusion coefficients of CO₂ are provided to evaluate the impact of amine functionalization of MOF-808 on the mass transfer of CO₂. Finally, the recoverability of the adsorbent is assessed by consecutive cycles to ascertain its stability for further applications.

Experimental

Materials

High pure Graphite powder (> 99.8%), potassium permanganate (KMnO₄, 99%), sulfuric acid (H₂SO₄, 98%), phosphoric acid (H₃PO₄, 85%), Hydrochloric acid (HCl, 37%), Methanol (HPLC grade), and Hydrogen peroxide (H₂O₂, 30%) were supplied from Merck company and used to prepare graphene oxide solid sorbent. Also, raw materials such as 5-Aminoisophthalic acid (AIPA, > 98%), 1,3,5-Benzenetricarboxylic acid (BTC, > 98%), Zirconium (IV) oxide chloride octahydrate (ZrOCl₂·8H₂O, 98%), Formic acid (> 98%), N,N-Dimethylformamide (DMF, > 99.9%), Chloroform (> 99%), and Acetone (> 99.5%) were purchased from Merck company and applied as the precursor for synthesizing MOF-808 adsorbent. All of the chemicals mentioned above were used without any purification.

GO preparation

The GO nanoparticle was synthesized through the improved Hummer’s technique⁵⁰, reported previously by Kumar, et al.⁵¹. A similar procedure was applied as follows: the concentrated H₂SO₄ (180 ml) and concentrated H₃PO₄ (20 ml) were poured into a flask, followed by adding the graphite powder (1.5 g) to the resulting acidic solution. Then, the potassium permanganate powder (9 g) was progressively added to the flask while the temperature was maintained at 40 °C. After the KMnO₄ powder was fully added, the temperature was gained to 60 °C, and the resulting mixture was agitated for 12 h. The solution’s color changed from brown to yellow as a consequence of cooling the flask’s contents and adding the obtained mixture to an ice bath (200 ml) that contained concentrated hydrogen peroxide (2 ml). The next step was to filter the mixture, and the obtained solid sorbent was then purified several times by utilizing distilled water (200 ml), HCl solution (200 ml, 30%), and pure methanol (200 ml). Following the purification stage, the GO sample was produced by centrifuging the resultant powder for 45 min at 5500 rpm, followed by 12 h of drying in a vacuum oven at 65 °C. The general procedure of graphene oxide synthesis is illustrated in Fig. 1.

Figure 1

General procedure of graphene oxide synthesis.

MOF-NH₂/GO-x% composites synthesis

To synthesize the MOF-NH₂/GO composite, a solvothermal method was performed similar to the procedure conducted by Esfahani et al.⁴⁹. To prepare MOF-NH₂/GO-22.5% composite, 1,3,5-Benzenetricarboxylic acid (0.066 g), 5-Aminoisophthalic acid (0.014 g), and the GO (0.058 g) were dissolved in DMF (10 ml) which resulted in solution A. Solution B was prepared through dissolving ZrOCl₂ salt (0.12 g) in the formic acid (20 ml). Next, the resulting solutions were individually blended in an ultrasonic bath for 20 min to reaching complete homogeneity. Subsequently, both solutions were introduced into a flask and mixed in the ultrasonic bath for 20 min. Then, the resulting mixture was introduced into a Teflon lined autoclave with a volumetric capacity of 300 ml and heated at 130 °C in a Muffle furnace for 48 h. After 48 h, the excess solvent was extracted from resulting composite using a Pasteur pipette, and the obtained composite was subjected to three consecutive washes with DMF (30 ml for 5 min) followed by three consecutive washes with acetone (30 ml for 5 min). Finally, the resulting composite sample was subjected to centrifugation at a speed of 6000 rpm for 8 min. Following the process of washing and centrifugation, the precipitate was transferred into an oven set at 100 °C, allowing for complete drying. Following its separation, the dried precipitate was combined with chloroform and heated for 15 min until it was homogenous. It was then placed in a furnace set at 50 °C for 72 h. After eliminating the extra solvent, the residual MOF-NH₂/GO-22.5% composite was dried by heating it at the temperature of 100 °C in an oven. Figure 2 shows the MOF-NH₂/GO-22.5% composite’s synthesis procedure.

Figure 2

Mixed ligand MOF-NH₂-GO composite’s synthesis procedure.

MOF-NH₂/GO composite characterization

The resulting MOF-808/GO composite with the different structures and elemental composition were investigated using various characterization techniques. Fourier Transform Infrared spectroscopy (FTIR) analysis was done using (Spectrum Rx1, Perkin Elmer Company) apparatus in the wave number range between 400–4000 cm⁻¹ to detect the functional groups in the composite samples. The porous characteristics of the obtained composites were characterized by Brunauer–Emmett–Teller (BET) method through conducting nitrogen adsorption–desorption analysis using a Micromeritics device (Model ASAP 2020, USA) at the temperature of 77K. Also, the Barrett–Joyner–Halenda (BJH) method was applied to achieve the pore size distribution curves of the samples. The samples’ morphological characteristics were analyzed using Field emission scanning electron microscopy (FESEM) and transmission electron microscopy (TEM) analysis, and the elemental composition of the composites were characterized by performing energy dispersive spectroscopy (EDS) analysis by a S-4700 microscope instrument (Hitachi, Japan). The composites’ thermal stability were tested by performing thermogravimetric analysis (TGA) in the temperature range between 50 to 800°C under an argon (Ar) atmosphere using a TGA analyzer model STA449F3, NETZSCH, Germany.

CO₂ adsorption setup

An experimental volumetric setup, illustrated in Fig. 3, was used to evaluate the CO₂ uptake capabilities of the resulting MOF-808/GO composites. As shown in Fig. 3, the pressure of the CO₂ stream is first regulated using a regulator device installed on the CO₂ cylinder. Next, the stream is warmed up by passing inside an electrical heater, and then the stream is charged to a stainless steel chamber, namely a mixing chamber, aiming to make the stream’s pressure and temperature uniform. After this, the CO₂ gas enters into a stainless steel adsorption bed with an overall volume of 254 cm³, height of 9 cm, and diameter of 6 cm. Consequently, the gas keeps reaching the composite bulk powder and the CO₂ adsorption process starts. The adsorption bed’s temperature is tuned to a specific value according to the set point temperature utilizing a water jacket installed in the outer shell of the adsorption vessel. The adsorption bed’s features, such as temperature and pressure, are measured every second and saved in a computer device with their corresponding adsorption time. Considering the gradual decrease in the pressure of the CO₂ inside the vessel over time, the CO₂ uptake capability of the composite sample can be measured with the help of Eqs. (1) and (2).

$$q=\frac{{m}_{i}-{m}_{f}}{W}=\left(\frac{V\times {\text{Mw}}}{R\times W}\right)\times \left({\left[\frac{P}{Z\times {\text{T}}}\right]}_{i}-{\left[\frac{P}{Z\times {\text{T}}}\right]}_{f} \right)$$

(1)

$$Z=1+\frac{B\times P}{R\times T}$$

(2)

where m_i and m_f indicate the initial and final mass of the CO₂ over the adsorption time. Also, some parameters such as P refer to pressure, V refers to the adsorption bed’s overall volume, W refers to the adsorbent’s mass, T is temperature, M_w is the molecular weight of the adsorbed gas, R is the global gas constant, and Z is the CO₂ compressibility factor. The virial’s coefficient, abbreviated as term B, can be calculated with the help of the Tsonopoulos correlation⁵² reported in Eqs. (3)–(5).

Figure 3

Schematic of the CO₂ adsorption setup.

$$B=\frac{R\times {T}_{c}}{{P}_{c}} ({F}^{\left(0\right)}{(T}_{r})+\upomega {F}^{\left(1\right)}{(T}_{r}))$$

(3)

$${F}^{\left(0\right)}{(T}_{r})=0.1445-\frac{0.330}{{T}_{r}}-\frac{0.1385}{{T}_{r}^{2}}-\frac{0.0121}{{T}_{r}^{3}}-\frac{0.000607}{{T}_{r}^{8}}$$

(4)

$${F}^{\left(1\right)}{(T}_{r})=0.0637+\frac{0.331}{{T}_{r}^{2}}-\frac{0.423}{{T}_{r}^{3}}-\frac{0.008}{{T}_{r}^{8}}$$

(5)

Design of experiment by RSM

Recently, utilizing the design of experiment (DOE) has gained prominence across many engineering disciplines. DOE serves as a valuable tool for establishing relationships between independent factors and the corresponding response while accounting for potential interactions among these variables. The use of RSM allows for the investigation and modeling of scenarios in which the outcome of a model is affected by independent variables within a designated design area. Examining the relationship among variables regarding reducing the number of trials is seen as a primary benefit of RSM approach. The coefficients of a quadratic model, as expressed in Eq. (6), can be identified by fitting the experiment’s results to the model as mentioned above. Additionally, the significance of the resulting model is assessed through an analysis of variance (ANOVA), utilizing statistical indexes including the P-value, F-value, and correlation coefficient (R²) of the obtained model^3,53. The R² coefficient and the absolute average relative error (AARE%) can be computed employing Eqs. (7) and (8), respectively⁵⁴.

$$Y={\upbeta }_{0}+{\sum }_{i=1}^{n}{\upbeta }_{i }{x}_{i}+ {\sum }_{i=1}^{n}{\upbeta }_{ii }{x}_{ii}^{2}+ {\sum }_{\begin{array}{c}i=1\\ j>i\end{array}}^{n}{\upbeta }_{ij }{x}_{i}{x}_{j}$$

(6)

$$ R^{2} = \frac{{(q^{\exp } - \overline{q}^{cal} )^{2} }}{{\sum\limits_{i = 1}^{N} {\left( {(q^{\exp } - \overline{q}^{cal} )^{2} + (q^{\exp } - q^{cal} )^{2} } \right)} }} $$

(7)

$$ \% AARE = \left( {\sum\limits_{i = 1}^{N} {\left| {\frac{{q^{\exp } - q^{cal} }}{{q^{\exp } }}} \right|/N} } \right) \times 100 $$

(8)

Which the model’s parameters, including ${\upbeta }_{0}$,${\upbeta }_{i}$, ${\upbeta }_{ii}$, ${\upbeta }_{ij}$, and n indicate the model’s constant term, linear effect, quadratic effect, the model’s parameters interaction term, and the number of variables. Central Composite Design (CCD) approach of the RSM technique is well recognized as a practical approach for establishing the relationship between independent variables and the corresponding response. In this approach, the variables’ range is divided into five discretized intervals: − α, − 1, 0, + 1, and + α which α value is correspond to the axial or star points in the orthogonal CCD design space⁵⁵. This study employed the Design of Experiments (DOE) methodology using the Design Expert software (version 11). The RSM-CCD technique was utilized to establish the relationship between the variables such as process’s temperature (T), process’s pressure (P), and the GO’s loading (wt.%) in the MOF-808/GO composite sample. A summary information about the variables’ name, abbreviation and their levels in the RSM-CCD method respect to considering the α value equal to 2 are reported in Table 1.

Table 1 Variables intervals in the RSM-CCD method.

Results and discussion

Adsorbents characterization

The FTIR spectrum of the MOF-NH₂/GO-x% (x is the weight percent of the GO in the composite sample) composites are exhibited in Fig. 4a. The spectral analysis of the composites revealed the presence of representative peaks at 659 and 757 cm⁻¹, which may be ascribed to the Zr–O–Zr bond’s stretching vibration and the Zr–OH bond’s vibration, respectively. Additionally, the bands seen at 1382 and 1448 cm⁻¹ are associated with the carboxylate group’s asymmetric stretching vibration. The peaks that appeared at 1584 and 1643 cm⁻¹ are attributed to the symmetric stretching vibration of the carboxylate group⁵⁶. In all samples, the bending and scissoring vibrations of N–H occurred at 1430 and 1575 cm⁻¹, respectively, suggesting that the amino group is present in these samples. Also, a broad peak around 3434 cm⁻¹ is related to the stretching vibration of the O–H bond in the hydroxyl or carboxyl group⁵⁷. The N₂ adsorption–desorption isotherms, illustrated in Fig. 4b, were prepared to evaluate the porosity characteristics of the synthesized composites. Based on the findings presented in Fig. 4c, the notable rise in N₂ adsorption observed at a relative pressure (P/P⁰) below 0.05 can be attributed to the existence of micropores within the structure of the MOF-NH₂/GO composite samples. Additionally, the presence of a distinct hysteresis loop at higher relative pressures (P/P⁰ > 0.8) may be indicative of the presence of larger pores or intraparticle channels within the solid sorbent’s skeleton⁵⁸. This observation aligns with the typical type IV isotherm classified by the international union of pure and applied chemistry (IUPAC)⁵⁹. Table 2 presents a summary of the surface area and porous characteristics of the obtained composite samples. Based on the results shown in Table 2, it can be observed that the BET surface area of the composites decreases with increasing GO content. The decrease, in the surface area of the composites can be attributed to the presence of oxygen containing groups on GO that compete with the ligands of MOF for coordination with metallic centers resulting in a distortion of the MOFs structure. This distortion affects both the distribution of sizes and the crystallization process of MOF. Furthermore incorporating GO appears to create regions between GO and MOF units thereby increasing volume in these composites^60,61. The results of the TGA analysis of the all composite samples are exhibited in Fig. 4d. It is clear that weight losing of the samples can be observed in three region namely region I, region II, and region III. In the region I (temperature lower than 200 °C) about 6.2% of samples’ weight are decreased which can be related to the vaporization of the adsorbed water or other solvents⁶². In the region II (temperature between 200 to 400 °C), decrease in the samples’ weight may be attributed to the removal of the modulators, linkers, and dehydroxylation of the zirconium nodes. In region III (temperature above 400 °C), thermal decomposition of functional groups containing oxygen and nitrogen, such as amine group, carboxylic acid and epoxy, causes a drastic reduction in the weight of composites, so that the remaining solid weight is equal to 46.49%, 49.96%, 52.71%, 55.92%, and 58.58% for composite samples including MOF-NH₂/GO-0%, MOF-NH₂/GO-7.5%, MOF-NH₂/GO-15%, MOF-NH₂/GO-22.5%, and MOF-NH₂/GO-30%, respectively. Increasing in the weight residual of the composite samples with a higher amount of GO may be related to the increasing in the weight percentage of the graphitic layers in the composite sample that remains unoxidized in the argon atmosphere⁶³.

Figure 4

(a) The FTIR spectra of the MOF-NH₂/GO-x% composite samples, (b) the results of the N₂ adsorption–desorption at 77.3 K, (c) pore size distribution plots of the MOF-NH₂/GO-x% samples, (d) TGA analysis results of resulting composites with the various GO’s loading amount.

Table 2 Porous properties of the obtained MOF-NH₂/GO-x% composites.

The scanning electron microscopy (SEM) images of the MOF-NH₂/GO-x% (the x symbol refers to the 0, 7.5, 15, 22.5, and 30%) are displayed in Fig. 5a–e, respectively. The findings suggest that the MOF-NH₂ sample’s structure (Fig. 5a) exhibited the formation of octahedral microcrystals characterized by a very homogeneous and evenly distributed particle arrangement. Figure 5b–e exhibit successful nucleation of the MOF/NH₂ on the surface of the GO, which corresponded with the results reported in the literature^37,64. The resulting composite samples’ particle sizes were measured with the help of transmission electron microscopy (TEM) analysis. The results were plotted as particle size distribution plots, as shown in Fig. 5f. Based on the findings of this figure, the pristine MOF sample (MOF-NH₂/GO-0%) exhibited a narrower particle size than other samples. Also, considering these plots, it can be concluded that increasing the GO quantity during the composite’s synthesis procedure causes a gradual increase in the resulting composite adsorbent’s particle size. Particle sizes in composites increase for a number of reasons. The extra GO layers provide more nucleation sites, which aids in the formation of MOF crystals. Functional groups on GO exhibit a significant affinity for metal ions, facilitating the process of nucleation and growth^65,66. The hierarchical arrangement of GO serves as a framework, augmenting the durability and dimensions of the MOF particles. The synergy of this mutually beneficial interaction and the reinforcement provided by the structure results in larger particle sizes in the composites⁶⁷. According to the results of the TEM analysis, the average particle sizes for the adsorbents namely MOF-NH₂/GO-x% (the x symbol refers to the 0, 7.5, 15, 22.5, and 30%) were obtained around 16.32, 24.30, 30.52, 43.52, and 55.97 nm, respectively. The energy dispersive spectroscopy (EDS) analysis was applied to assess the elemental composition of the obtained composite samples and the results are summarized in Table 3. Based on the findings of the EDS examination, it can be determined that the synthesized samples primarily consist of the elements carbon (C), oxygen (O), nitrogen (N), and zirconium (Zr). According to the results of Table 3, a gradual increase in the carbon and oxygen elements can be observed in the composite samples with a higher amount of the GO. It can be attributed to the oxygen rich functional groups such as carboxylic acid, hydroxyl, and epoxy incorporated into the surface of the GO.

Figure 5

SEM image of the MOF-NH₂/GO-x% composite adsorbents: (a) x = 0%, (b) x = 7.5%, (c) x = 15%, (d) x = 22.5%, and (e) x = 30%. (f) Particle size distribution plot of the resulting MOF-NH₂/GO-x% composites obtained from TEM analysis.

Table 3 Elemental composition of the MOF-NH₂/GO-x% samples based on EDS analysis.

RSM results

Analysis of variance (ANOVA)

Table 4 presents the ANOVA findings for the process of CO₂ adsorption utilizing MOF-NH₂/GO-x% composites, as well as the effect of several parameters, including the temperature and pressure of the adsorption process, as well as the quantity of GO loading on the obtained composite sample. In order to assess the precision of the RSM-based model acquired inside the design space, two statistical criteria are introduced: the P-value and the F-value. The P-value indicates the significance of the produced model or its terms, with a threshold of 0.05. Conversely, the F-value should exceed 1 to demonstrate the significance of the model or its terms. Hence, based on the analysis of the P-value (less than 0.0001) presented in Table 4, it can be inferred that the developed model exhibits sufficient accuracy within the designated experimental range. Furthermore, the P-values associated with the model’s variables indicate that the factors of temperature and pressure hold significant importance and exert a greater influence than the GO loading (%) variable in the established model. The F-value of the model (129.07) demonstrates a considerable increase compared to the baseline value of 1, indicating the model’s significance and the minimal possibility of it being a random occurrence (probability < 0.01%)⁶⁸. The reliability of the created model can be established based on the statistical measures. The R² value of 0.991 above the threshold of 0.8, indicating a high level of explanatory power. Additionally, the discrepancy between the predicted R² value of 0.934 and the adjusted R² value of 0.983 is within the acceptable range of 0.2. The predictive model that has been built exhibits potential use in the design of industrial processes, as it demonstrates an adequate precision value of 43.29 for signal-to-noise evaluation, above the threshold of 4.

Table 4 ANOVA findings of the CO₂ adsorption process modeling based on RSM-CCD.

CO₂ adsorption RSM-CCD correlation

A quadratic model is proposed to investigate the relationship between the aforementioned parameters, namely the temperature of the adsorption process, the pressure, and the GO loading quantity (%), with regard to the CO₂ uptake capability of the MOF-NH₂/GO-x% composites. This model is derived from the RSM technique using the CCD approach and is represented by Eq. (9). Furthermore, Fig. 6 depicts the comparison of the predicted CO₂ adsorption amounts obtained from the RSM-based model with the corresponding experimental values. The figure demonstrates that the semi-empirical model exhibits a higher accuracy level due to the close alignment of the predicted and experimental values of CO₂ adsorption capacity along the diagonal line, indicating a suitable level of centralization.

Figure 6

Predicted vs. experimental amount of CO₂ adsorption capacity.

$${\text{q}}_{{\text{CO}}_{2}}= -2.49848+0.067353\times \text{A}+50.69443\times \text{B}+4.32582\times \text{C}-0.391931\times \text{AB}-0.054022\times \text{AC}+0.574425\times \text{BC}+0.008606\times {\text{A}}^{2}-2.04489\times {\text{B}}^{2}-0.173250\times {\text{C}}^{2}$$

(9)

Effect of operating conditions on CO₂ adsorption

In order to examine the impact of input parameters on the CO₂ adsorption capabilities of the MOF-NH₂/GO-x% composites, a perturbation plot was generated and is presented in Fig. 7. The curves labeled A, B, and C in the diagram represent the variables of temperature, pressure, and GO loading amount, respectively. Based on the outcomes presented in Fig. 7a, it is evident that the MOF-NH₂/GO-x% adsorbent exhibits an enhanced capacity for CO₂ uptake as the loading of GO in the composite sample is increased up to 15.2%. However, beyond this threshold, further increases in the GO loading result in a gradual reduction in the adsorption capacity of the solid sorbent for CO₂. The enhanced CO₂ adsorption performance of the MOF-NH₂/GO composite may be attributed to the increased surface potential energy of the composite due to the incorporation of electron-withdrawing functional groups, including carboxylic acid, hydroxyl, and epoxy, within the structure of the GO sample. The dispersion of the mentioned functional groups leads to an increase in the heterogeneity of the composite sample. Therefore, it enhances the CO₂ uptake capability of the MOF-NH₂/GO composite by improving the dipole-quadrupole interaction between the surface of the adsorbent and CO₂ molecules⁶⁹. Nevertheless, exceeding the threshold loading value of the GO (more than 15.1%) leads to a reduction in the composite sample’s capacity to adsorb CO₂. The phenomenon described can be attributed to pore filling, which reduces the pore width in the adsorbent and the subsequent clogging of pores. This effect is observed particularly at high loading levels of the GO⁴¹. As seen in Fig. 7a, the augmentation of pressure in the adsorption process leads to a significant improvement in the CO₂ adsorption capacity of the MOF-NH₂/GO-x% composite. The observed phenomenon may be ascribed to the enhanced mass transfer of CO₂ molecules into the cavities that were previously inaccessible, along with a reduction in the desorption of the trapped CO₂ molecules³⁷. Based on the observed trend of decreasing CO₂ adsorption with increasing process temperature, it can be inferred that the reduction in CO₂ uptake may be attributed to the dominant physical adsorption of CO₂ molecules on the surface of the composite sample. Additionally, the elevated temperature enhances the mobility of the absorbed CO₂ molecules, thereby facilitating their desorption at higher temperatures⁷⁰.

Figure 7

Effect of the parameters on the CO₂ uptake capability of the MOF-NH₂/GO-x% composite: (a) all factors, (b) the GO loading amount and the pressure at T = 45 °C, (c) the GO loading amount and the temperature at P = 5 bar, and (d) temperature and pressure when GO = 15%.

In order to effectively demonstrate the impact of the parameters mentioned above and their interaction on the CO₂ adsorption capacity of the MOF-NH₂/GO-x% composite samples, three-dimensional graphs were generated that are presented in Fig. 7c,d. Based on the Fig. 7b, it can be observed that increasing in the quantity of the GO loading in the MOF-NH₂/GO composite causes a gradually decreasing in the CO₂ uptake capability of the composite at the lower pressure while at the higher pressure, improving in the GO’s loading up to 19.6% gains the CO₂ uptake capability of the adsorbent. Therefore, it can be concluded that the pore blocking effect of the GO can be more significant in the CO₂ adsorption process by using MOF-NH₂/GO adsorbent at the ambient pressure. Figure 7c represents the dependency of the CO₂ adsorption to the temperature and the GO’s loading amount at the constant pressure of 5 bar. According to this figure, it can be resulted in increasing in the adsorption process’s temperature causes decreasing in the CO₂ adsorption capacity of the composites, but it should be considered that at the higher temperature (T = 65 °C) the MOF-NH₂/GO sample with the higher GO’s loading amount (GO = 30%) exhibited lower CO₂ uptake capability in comparison with the MOF-NH₂/GO sample with the lower GO’s loading quantity (GO = 0%) at the same temperature. This phenomenon can be related to the lower density of the amine group in the composite sample with the highest GO’s loading in comparison to the MOF-NH₂ sample (GO = 0%), as well as, the more favorability of the CO₂ chemisorption by amine active sites at the higher temperature⁷¹. Based on the presented data in the Fig. 7d, it can be shown that the maximum pressure and the minimum process’s temperature can be considered as the most favorable operational condition which under this condition the MOF-NH₂/GO-15% composite exhibited the greatest capacity for CO₂ adsorption.

Optimization of CO₂ adsorption process

The optimization of the operating parameters is crucial in industrial applications of the MOF-NH₂/GO-x% composite in order to maximize its capacity for CO₂ adsorption. The optimization of CO₂ adsorption was conducted using the optimization module of the Design Expert software. In order to ascertain the maximum uptake capacity of the composite, the effective parameters, namely GO loading quantity, temperature, and pressure, were established within a specified range. The objective was to optimize the adsorption capacity, and an optimization process was then conducted. The maximum desirability is seen under the optimal situation, as shown in Table 5. The experimental verification of the optimized CO₂ adsorption capability of the RSM-CCD technique was conducted by performing the CO₂ capture experiment at the obtained optimal condition. After conducting five replications of the adsorption test under optimal circumstances, a negligible AARD of around 1.42% was observed.

Table 5 The optimal operational condition of CO₂ adsorption based on the RSM-CCD.

Adsorption isotherms

Isotherm modeling may be a valuable tool in the design of adsorption systems since it can offer pertinent data on the mechanism of the adsorption process. These models elucidate the underlying mechanisms and dynamics involved in the interactions and reactions between the surface of solid sorbents and the adsorbate molecules during gas capture processes. The present study used isotherm models, including Langmuir, Freundlich, Dubbin–Radushkevich (D–R), Temkin, and Sips as two-parameter models for elucidating the adsorption phenomenon. The models indicated above are denoted as Eqs. (10)–(14), respectively⁷².

$$ {\text{Langmuir:}}\quad q_{e} = \frac{{q_{m} K_{l} P_{e} }}{{1 + K_{l} P_{e} }} $$

(10)

$$ {\text{Freundlich}}:\quad q_{e} = k_{F} P_{e}^{{{\raise0.7ex\hbox{$1$} \!\mathord{\left/ {\vphantom {1 n}}\right.\kern-0pt} \!\lower0.7ex\hbox{$n$}}}} $$

(11)

$$ {\text{Dubinin}} - {\text{Radushkevich}}:\quad q_{e} = q_{s} {\text{exp}}\left( { - k_{ad} \times E_{a}^{2} } \right) $$

(12)

$$ {\text{Temkin:}}\quad q_{e} = B Ln A + B Ln P_{e} ,\;\;B = \left( {\frac{RT}{{b_{T} }}} \right) $$

(13)

$$ {\text{Sips}}:\quad q_{e} = \frac{{ q_{s} ( \beta_{s} P_{e} )^{{ 1/a_{s} }} }}{{1 + ( \beta_{s} P_{e} )^{{ 1/a_{s} }} }} $$

(14)

The variables q_e and P_e represent the equilibrium CO₂ adsorption capacity (mg·g⁻¹) and equilibrium pressure (bar). Other parameters of the models, such as q_l, represent the maximal adsorption capacity (mg·g⁻¹) based on the Langmuir model. K_l (bar⁻¹) is the constant of the Langmuir model. In the Freundlich isotherm model, the terms including K_F (mg·g⁻¹·bar^1/n) and n_F are the model’s constants. In the Temkin model, the constant A_T (expressed in L·mol⁻¹) represents a fundamental parameter, while B denotes the first coefficient of the virial equation of state ($B=\left(\frac{RT}{{b}_{T}}\right), {b}_{T}=(J.{mol}^{-1})$). In the D–R model, the terms λ (mol²·j⁻¹) and ω (j·mol⁻¹) correspond to the model’s constant and the Polanyi potential, respectively. Additionally, within the Sips model, the terms q_s (mg·g⁻¹) is the maximum uptake capacity, k_s (bar⁻¹) is the model’s constant, and the term α_s is a factor that characterizes the heterogeneity of the system, which may be caused by the solid, the adsorbate, or a combination of the two^73,74.

In order to conduct isotherm modeling, a new composite sample with the optimal GO’s loading quantity was fabricated (MOF-NH₂/GO-22.6%), and the CO₂ adsorption experiments were performed under varying pressures ranging from 1 to 9 bar and at temperatures of 298, 308, 318, 328, and 338 K. Subsequently, the models as mentioned earlier were applied to the equilibrium adsorption data, and the corresponding parameters of each model were determined. These parameters are shown in Table 6. Additionally, Fig. 8 illustrates the fitted curves of the isotherm models at a temperature of 298 K and 318 K. The data from Table 6 demonstrate that Langmuir model’s constant (q_l) follows a falling order as temperature increases, providing evidence for the exothermic character of the CO₂ adsorption process. Furthermore, the K_F value of the Freundlich model, which represents the propensity of the adsorbate molecule to be adsorbed by the solid sorbent, exhibits a gradual decrease with increasing temperature. This phenomenon suggests that the adsorption of CO₂ by the MOF-NH₂/GO-22.6% sample is dominantly driven by physisorption mechanisms⁷⁵. Also, a favorable CO₂ uptake process is indicated by Freundlich model’s parameter (n_F), which ranges between 1 and 2. Hence, it may be inferred that the surface of the obtained composite exhibits heterogeneity, and the adsorption of CO₂ occurs in multilayers on the surface of the adsorbent. In addition, the constant (ω) in the D–R model and the term B in the Temkin model represent the free energy of adsorption, the ω and the B values below 8 kJ/mol suggest a physisorption process, but ω values ranging from 8–16 kJ/mol imply chemisorption. The ω values below 8 kJ/mol provide more evidence supporting the previously reported phenomenon of physical adsorption of CO₂ molecules⁷⁶. The term α_s in the Sips model is the heterogeneity factor that represents a divergence from the linearity of adsorption. The values of α_s are more than one in examined temperatures, indicating that the adsorption system becomes more heterogeneous by increasing temperature⁷³. The following is the order of the isotherm models’ abilities to forecast and characterize the adsorption behavior while taking the correlation coefficient (R²) criterion into account: Sips > Langmuir > Freundlich > Temkin > Dubinin–Radushkevich.

Table 6 Isotherm models’ characteristics at the temperature of 298, 308, 318, 328, and 338 K.

Figure 8

CO₂ uptake isotherm models fitted curves at the temperature of (a) 25 °C and (b) 45 °C.

Adsorption kinetics

In general, examining the adsorption kinetics can yield valuable insights into the rate of adsorption, which substantially impacts the effectiveness of adsorption. Kinetic models are commonly utilized to ascertain the retention period of the adsorbate molecules in the adsorption bed and the rate of their uptake, particularly in the context of industrial process design purposes like fixed bed columns. Hence, to examine the kinetics of CO₂ uptake in this study, some commonly used kinetic models, including first-order, second-order, Elovich, Rate Controlling, and Fractional order models, were employed. The kinetic models discussed above are shown in Eqs. (15)–(19), respectively. The CO₂ adsorption data were obtained by experimentation using the recently developed MOF-NH₂/GO-22.6% composite sample, obtained previously as the most appropriate GO loading quantity with the highest CO₂ adsorption capacity. The experiments were conducted at a pressure of 5 bar and temperatures of 298, 308, 318, 328, and 338 K. Table 7 provides a summary of the outcomes obtained by applying kinetic modeling, including the corresponding model parameters. Additionally, Fig. 9 depicts the graphical representation of the kinetic models through plotted curves at the pressure of 5 bar and the temperatures of 298, 308, 318, and 328 K.

$$ {\text{First order}}:\quad q_{t} = q_{e} \left( {1 - e^{{k_{1} t}} } \right) $$

(15)

$$ {\text{Second order:}}\quad q_{t} = \left( {q_{e}^{2} k_{2} t} \right)/\left( { 1 + q_{e} k_{2} t} \right) $$

(16)

$$ {\text{Elovich:}}\quad q_{t} = \beta \ln \left( {\alpha \beta } \right) + \beta {\text{ln}}\left( t \right) $$

(17)

$$ {\text{Rate controlling:}}\quad q_{t} = k_{c} t^{0.5} $$

(18)

$$ {\text{Fractional order:}}\quad q_{t} = q_{e} - \left[ {\frac{{\left( {n - 1} \right)}}{m} k_{n} t^{m} + q_{e}^{{\left( {1 - n} \right)}} } \right]^{{\left( {1/\left( {1 - n} \right)} \right)}} $$

(19)

Table 7 The characteristics of the kinetic models at the different temperatures and the pressure of 5 bar.

Figure 9

Fitted plots of the kinetic models at the pressure of 5 bar and the temperature of (a) 298 K, (b) 308 K, (c) 318 K, and (d) 328 K.

The constants k₁, k₂, k_c, and k_n represent the parameters of their respective kinetic models^77,78.

The rate of CO₂ adsorption by composites significantly rises as a result of various factors. Zr-MOFs possess a large surface area and porosity, which enables rapid diffusion and attachment of CO₂ molecules. This is due to the presence of numerous active sites and the easy accessibility to internal surfaces⁷⁹. Zirconium centers possess Lewis acid sites that efficiently interact with electron-rich CO₂ molecules, leading to enhanced adsorption kinetics. Linkers that have been modified with functional groups such as amino or hydroxyl improve Van der Waals interactions, thereby increasing both the capacity and speed of adsorption^80,81. The first-order model operates on the idea that the adsorption rate is directly proportional to the discrepancy between the saturation concentration and the quantity of the adsorbed species on the adsorbent’s surface. The diminishing impact of chemical adsorption on the adsorption process is evident in the decrease of the R² value in the first-order model, as reported in Table 7⁸². According to the Rate Controlling Model, the adsorption rate is influenced by intraparticle diffusion. The model’s R² value exhibits an upward trend as temperature increases, indicating that diffusion may serve as the dominant mechanism governing the rate of the process⁸³. Table 7 demonstrates the efficacy of the fractional order kinetic model in establishing a correlation between the adsorption capacity of CO₂ and the adsorption time, as indicated by the R² value of the latter model. The latter model offers a more precise depiction of the adsorption process, deviating from the kinetics of integer order. The complexity of the adsorption process is influenced by several parameters, including surface heterogeneity, multilayer adsorption, and interactions between adsorbate molecules and the adsorbent surface⁸⁴.

Adsorption thermodynamic

Thermodynamic analysis was used to calculate parameters like entropy changes (ΔS), Gibbs free energy changes (ΔG), and enthalpy changes (ΔH) in the CO₂ adsorption process. Distribution factor (K_d) under constant pressure of 5 bar and the temperatures of 298, 308, 318, and 328 K was computed using Eq. (20). The Vant Hoff plot (Fig. 10) was created by charting the K_d values that were acquired against the corresponding reverse temperature. The van’t Hoff plot’s slope and intercept, as shown in Eq. (21), correspond to the entropy change (${{\Delta {\text{S}}}}^{0}$) and enthalpy change ($\Delta {\text{H}}^{0}$), respectively. Equation (22), which measures the process’s ($\Delta {\text{G}}^{0}$), can be utilized. Table 8 reports the thermodynamic characteristics of the CO₂ adsorption process.

$${K}_{d}=\Delta P\times \frac{V}{w}$$

(20)

$$\text{ln}\left({K}_{d}\right)=\frac{\Delta {\text{S}}^{0}}{R}- \frac{\Delta {\text{H}}^{0}}{RT}$$

(21)

$${\Delta \text{G}}^{0}={\Delta \text{H}}^{0}-T{\Delta \text{S}}^{0}$$

(22)

where V is the volume of the gas inside the vessel, W is the mass of the solid sorbent, and R is the gas constant (8.314 J·mol⁻¹·K⁻¹)⁸⁵. The pressure differential in the adsorption vessel during the CO₂ uptake process is represented by ∆P. Based on the data documented in Table 8, the ΔH⁰ value of − 16.905 kJ·mol⁻¹ indicates a dominant physically adsorption of the CO₂ molecules on the MOF-NH₂/GO-22.6% adsorbent. This value is in the range of heat releases below 20 kJ·mol⁻¹, which is often associated with physisorption mechanisms. In contrast, chemisorption mechanisms are characterized by heat releases⁸⁶ over 40 kJ·mol⁻¹. Furthermore, valuable insights into the randomized or organized interaction between the gas phase and the solid sorbent may be obtained from the adsorption ΔS⁰ value. A positive change in entropy ΔS⁰ indicates that the adsorption process is occurring in a more random manner, whereas a negative ΔS⁰ value suggests that the adsorption is less randomized. Based on the observed negative value of ΔS⁰ (− 0.030 kJ·mol⁻¹·K⁻¹), it may be deduced that the gas–solid interface exhibits a lower degree of randomness. The presence of negative values for the standard Gibbs free energy change ΔG⁰ at various temperatures indicates the favorable nature of the CO₂ capture process, as well as the spontaneous occurrence of the process⁸⁷.

Figure 10

Van’s Hoff plots of the CO₂ adsorption process at the pressure of 5 bar.

Table 8 The results of the thermodynamic analysis of the CO₂ uptake process at pressure of 5 bar.

CO₂ diffusion coefficient calculation

To show the dependency of the CO₂ molecules diffusion to the adsorbent’s particle size, the CO₂ diffusion coefficients were calculated, similar to the study conducted by Zhao et al.⁸⁸. The diffusion coefficients for all MOF-NH₂/GO-x% (x = 0, 7.5, 15, 22.5, and 30) composite samples were determined using the utilization of Eq. (23), as depicted below:

$${q_{t}}/{q_{e}}=\frac{6}{{r}_{c}}\sqrt{\frac{{D}_{m} t}{\pi }}$$

(23)

The uptake capacity at the time of t, equilibrium uptake capacity, the average radius of the composite’s particles, and time are denoted by the variables qt, qe, rc, and t, respectively. The mean of the diffusion coefficient within the adsorption time of 0 to t is also denoted by the term D_m⁸⁸. The CO₂ uptake capacity ratio (${{q}_{t}}/{{q}_{e}}$) for all samples are plotted against the square root of the adsorption time (t^0.5) in Fig. 11a. By fitting the adsorption capacity ratio data on a linear equation and utilizing the fitted line’s slope, the D_m value could be computed using Eq. (24). The results of the diffusion coefficient calculation are provided in Table 9.

Figure 11

(a) Plots of CO₂ uptake capacities ratio against the square root of time (t^0.5), (b) FTIR spectra of the MOF-NH₂/GO-22.6% sample before/after uptake process.

Table 9 The results of the CO₂ diffusion coefficient (D_m) calculation.

$${D}_{m}=\frac{{slope}^{2} {r}_{c}^{2} \pi }{36}$$

(24)

Table 9 indicates that increasing the MOF-NH₂/GO composite’s particle size increases the average diffusion coefficient (D_m), from 1.22 × 10⁻¹³ (cm²/s) for MOF-NH₂/GO-0% to 10.89 × 10⁻¹³ (cm²/s) for MOF-NH₂/GO-30% adsorbent. Consequently, it can be deduced that the greater diffusion coefficient value associated with the MOF-NH₂/GO-30% sample is due to its larger particle size, which results in more intraparticle voids inside the composite’s solid powder and lowers the CO₂ mass transfer barrier⁸⁸.

CO₂ adsorption mechanism

In order to study the mechanism of the CO₂ adsorption using MOF-NH₂/GO-22.6% composite, the FTIR spectra of the adsorbent was provided before/after the CO₂ uptake process. According to the outcomes of the FTIR spectra, illustrated in Fig. 11b, it can be concluded that CO₂ molecules are concurrently captured by chemically and physically adsorption pathways. The prominent peak at 2345 cm⁻¹ indicates the CO₂ stretching vibration, supporting the physisorption pathway. Meanwhile, new bands seen after CO₂ adsorption indicate the CO₂ molecules being chemically adsorbed by the composite sample. The peaks at 2998 cm⁻¹ and 1628 cm⁻¹ are associated with ammonium formation, in particular the ${\text{R}{-}\text{NH}}_{3}^{+}$ and ${\text{R}{-}\text{NH}}_{2}^{+}$ stretching vibration. Also, the formation of carbamate ions can be concluded based on the peaks observed at 1534 cm⁻¹ and 1689 cm⁻¹, which are attributed to the asymmetric and symmetric stretching vibrations of the COO⁻ species^89,90,91.

In general, the CO₂ molecules adsorption by the amine functional group is performed in a two-step reaction: first, the CO₂ molecules are captured by the primary amine through the production of a zwitterion intermediate ($\text{R}{-}{\text{NH}}_{2}^{+}\cdots {\text{COO}}^{-}$). Subsequently, the zwitterion intermediate undergoes deprotonation with the adjacent amine group, leading to the creation of ion pairs consisting of ammonium-carbamate (($\text{R}{-}{\text{NH}}_{3}^{+}\cdots {\text{COO}}^{-}{-}\text{NH}{-}\text{R}$)). Additionally, proton transfer might contribute to the formation of carbamic acid species ($\text{R}{-}\text{NH}{-}\text{ COOH}$)⁹¹. Based on the findings, the MOF-NH₂/GO-22.6% sample’s FTIR spectra validate the creation of the species as mentioned above following the CO₂ adsorption procedure.

NH₂/GO-22.6% composite recyclability

The solid adsorbent’s reusability should be examined as the most crucial economic factors for the development of novel solid adsorbents in order to assess the feasibility of CO₂ adsorption on an industrial scale. In order to examine the robustness and regenerability of the developed composite, a series of fifteen cycles of the CO₂ capture process were performed at a pressure of 5 bar and a temperature of 45 °C. During each cycle, the used adsorbent was recycled under a temperature of 95 °C in a vacuum oven for 2 h. The performance of the composite material in adsorbing CO₂ after 15 cycles is depicted and visualized in Fig. 12. Based on the data presented in the figure, a decrease of approximately 5.79% in the composite’s efficiency for absorbing CO₂ may be detected after 15 cycles. As a result, the composite that was produced is an effective choice for CO₂ uptake applications.

Figure 12

MOF-NH₂/GO-22.6% adsorbent regeneration efficiency after fifteen cycle.

Comparison of absorption capacity of MOF/GO composites

This section presents a comparative analysis of the CO₂ uptake capability of the MOF-NH₂/GO-22.6% composite in comparison to previous research conducted in the same field. Table 10 presents the findings of the investigations conducted on the adsorption of CO₂ using several types of MOFs and their composites with GO, together with the respective operating parameters. Based on the data reported in the table, it is evident that the synthesized composite of MOF-NH₂/GO-22.6% demonstrates a notably elevated capacity for CO₂ adsorption in a similar condition when compared to other adsorbents. Consequently, this composite holds significant potential as an efficient adsorbent for large-scale CO₂ capture applications in large-scale CO₂ uptake plants.

Table 10 Results of similar studies on the CO₂ adsorption using MOF/GO composites.

Conclusion

In summary, several mixed ligand/graphene oxide composites, namely MOF-NH₂/GO-x% were made with varying loading weight percent of GO, and CO₂ uptake experiments were carried out utilizing the resultant solid adsorbents. The RSM-CCD approach was used to examine the impact of several factors on the adsorbents’ capacity to adsorb CO_2, such as GO loading, temperature, and pressure. Utilizing the RSM method, perturbation plots and three-dimensional response surfaces were provided. These showed that increasing the GO’s loading up to 22.6 wt% improved the composite adsorbent’s capacity to adsorb CO₂, while increasing the amount of GO loading above the mentioned quantity decreased CO₂ adsorption capacity because the excess GO obstructed the pore. Following optimization using the RSM-BBD technique, the ideal MOF-NH₂/GO composite structure was discovered, with GO loading equal to 22.6 wt%. BET, FTIR, EDS, TEM, and SEM analysis were used to characterize the obtained composites’ structural and elemental characteristics. Although a decreasing trend in the composite samples’ BET surface area from 2021.37 to 1270.60 m²/g was observed via increasing in GO’s loading value from 0 to 22.6%, the results indicated that the CO₂ uptake capability of the samples was enhanced dramatically from 206.99 to 302.58 mg/g. This phenomenon originated from increasing the surface heterogeneity of the composite and enhancing dipole-quadrupole moment interaction between CO₂ molecules and the adsorbent’s surface due to the oxygen-rich nature of the GO. The isotherm modeling outcomes showed that the CO₂ adsorption process involved multilayers and that the surface of the MOF-NH₂/GO composite is heterogeneous. Furthermore, the best fit of the experimental CO₂ uptake data by the fractional order kinetic model demonstrated that the reaction order could not be an integer and that the adsorption kinetic was affected by a variety of variables, including heterogeneity and adsorbate-adsorbent interaction. Ultimately, the investigation of the adsorption mechanism exhibited both physisorption and chemisorption pathways of CO₂ adsorption, while the thermodynamic analysis revealed the predominant role of the exothermic physisorption process. After fifteen cycles of adsorption-desorption, the regeneration study revealed a high degree of stability and resilience of the adsorbent with respect to the slight losses in the CO₂ adsorption efficiency (around 5.79%).