Carbon Dioxide (CO2) Adsorption by Activated Carbon Functionalized with Deep Eutectic Solvent (DES)

In recent years, carbon dioxide (CO2) emission has become a major concern as the amount of the emitted gas significantly increases annually. Consequently, this phenomenon contributes to global warming. Several CO2 capture methods, including chemical adsorption by activated carbon, have been proposed. In this study, activated carbon was prepared from sea mango (Cerbera odollam), which was functionalized with deep eutectic solvent (DES) composed of choline chloride and glycerol to increase the efficiency of CO2 capture. The samples underwent pre-carbonization and carbonization processes at 200 °C and 500 °C, respectively, with nitrogen gas and flowing several gases, namely, CO2 and steam, and then followed by impregnation with 50 phosphoric acid (H3PO4) at 1:2 precursor-to-activant ratio. The prepared activated carbon was impregnated with DES at 1:2 precursor-to-activant ratio. The optimum CO2 adsorption capacity of the activated carbon was obtained by using CO2 gas treatment method (9.851 mgCO2/gsol), followed by the absence of gases (9.685 mgCO2/gsol), steam (9.636 mgCO2/gsol), and N2 (9.536 mgCO2/gsol).


Introduction
Current atmospheric carbon dioxide (CO2) levels are considered alarming; such levels have contributed to global environmental issues. Santos [1] reported that 70% of CO2 emission is attributed to different forms of energy generated and used in various processes, such as burning of fossil fuels, including natural gas, oil, or coal, and industrial processes, such as cement production, oil refinery, power plant processes, and iron and steel manufacturing. Furthermore, CO2 emission not only adversely affects climate change but also threatens human health.
Several effective methods of carbon capture and storage (CCS), such as adsorption [2], membrane separation [3], and cryogenic separation, have been proposed to reduce the amount of emitted CO2 in the atmosphere [4]. Lee [5] demonstrated that adsorption is a promising method to remove CO2 because it is cost effective.
Adsorption processes are generally performed using activated carbon [6], activated [7], zeolite [8], and polymeric adsorbents [9]. A few researchers [5,10,11] also revealed that activated carbon can capture CO2 because it consists of a large surface area per unit volume and submicroscopic pores, in which contaminant adsorption occurs. Moreover, activated carbon is stable under acidic and basic conditions. It is also cost effective because it can be regenerated and thus suitable for organic compound removal.
Considering cost effectiveness in activated carbon production, researchers developed different precursors from abundant waste materials, such as palm shells, sea mango, cocoa pod shells, and rice husks. For instance, Siti Noraishah [12] successfully produced activated carbon from rice husks and utilized it to remove CO2 generated from industrial activities.
The integration of chemicals into activated carbon has shown an excellent performance in CO2 uptake. Saad [13] and Karousos [14] demonstrated that the impregnation of activated carbon with chemicals, such as amine and ionic liquids (ILs), has increased the amount of captured CO2. Although the immobilization of these chemicals in activated carbon assists CO2 capture, such chemicals are difficult to prepare because they are toxic and relatively expensive [15]. Abbott [16] developed a novel green solvent known as Deep Eutectic Solvent (DES) to overcome the limitation of amine and ILs [17]. DES is used in numerous possible applications, including metal electro deposition, metal electropolishing, and synthesis applications, such as biodiesel purification and manufacturing, biotransformation, and CO2 capture [17]. DES for CO 2 capture has been extensively investigated because of its easy preparation, biodegradability, and nontoxicity; DES can also yield results similar to amine, which is rigidly known for good CO2 capture [18].
DES can be classified into four types: Type 1, quaternary ammonium cation + metal halide; Type 2, quaternary ammonium cation + hydrated metal halide; Type 3, quaternary ammonium cation + hydrogen bond donor; and Type 4, metal halide + hydrogen bond donor. Among these types, Type 3 is the most favorable because it is easily produced, mostly unreactive in water, biodegradable, and inexpensive [17].
In the current research, the prepared activated carbon from sea mango (Cerbera odollam) was modified to improve the CO2 adsorption capacity by promoting DES on the surface of the activated carbon through chemical impregnation. Thus far, very few studies have focused on CO2 capture with DES. Hence, the present work focused on the characterization and performance of CO2 capture by using DES-functionalized activated carbon.

Preparation of activated carbon
Sea mango (Cerbera odollam) was collected around Perlis and Kedah. The skin and seeds of the raw material were removed to obtain the fibrous shell. The samples were then washed and dried in an oven for 24 h at 105 °C to remove the moisture content. Afterward, the samples were crushed and sieved to obtain fine particles measuring 1 mm to 2 mm. The raw material was then subjected to semicarbonization in a tube furnace for 30 min at 200 °C by flowing with nitrogen gas. Char was subsequently impregnated with 50 wt% H3PO4 at 1:2 precursor-to-activant ratio for 4 h and then dehydrated in the oven at 110 °C until dryness.
The char was weighed and placed in a tube furnace with three different gases, namely, nitrogen, carbon dioxide, and steam, for activation. One tube furnace was not filled with gases. The furnace was heated until the desired temperature of 500 °C was reached. Activation was conducted for 2 h. The samples were cooled to room temperature (25°C) and washed few times with hot distilled water to remove excess H3PO4.
Activated carbon was then filtered and dried in the oven until it is completely dried. The activated carbon samples for gas activation of nitrogen, carbon dioxide, steam, and absence of any gases were denoted as AC-N2, AC-CO2, AC-S, and AC-AAG, respectively.

Preparation of DES
Choline chloride and glycerol (1:2) were added at 80 °C under stirring. The mixture was stirred for approximately 1 h, and the resulting solution was clear and homogeneous.

Functionalization of activated carbon with DES
Each type of activated carbon was impregnated by using DES with a 1:2 ratio of activated carbon to DES for 4 h by using the vacuum impregnation method with a vacuum pressure of −0.5 bar. The activated carbon was completely dried at 150 °C. The impregnated activated carbon samples for AC-N2, AC-CO2, AC-S, and AC-AAG were denoted as AC-N2-D, AC-CO2-D, AC-S-D, and AC-AAG-D, respectively.

Carbon dioxide adsorption
The experimental set-up employed in this study is as shown in figure 1. The adsorption column was made of glass with 1 cm inner diameter and 30 cm long, enclosed with an insulator to retain a constant temperature throughout the whole experiments. The glass column was vertically installed and implanted within a glass water jacket with inlet and outlet connections. The activated carbon (3 g) was inserted into this column with a piece of cotton wool at the top and bottom for breakthrough study. During the experiments, the gas flow was in an up-flow manner at 5ml/min.
A Guardian NG CO2 analyzer with measuring a range of 0-30% CO2 (Edinburgh Instruments Ltd, Chorley, Lancashire, England) was used to measure the concentration of CO2 exiting the adsorption bed. The CO2 analyzer displays a value when the adsorption column is saturated with CO2 and thus causes the gas to break through out of the column. The breakthrough time was recorded for every sample.
The gases flow from the gas cylinders was regulated using gas regulators Weldmark provided by (Weldmarks's Corporate, Indianapolis, Indiana, United State). A set of gas flow controller (Brooks Instrument, PA, USA) model SLA 5850 and model 0254 Secondary Electronics were used to accurately measure and control the gas flow rate. The gases, purified N2 and 15 % CO2 (balance N2) were obtained from The Linde Group (Munich, Germany).

Characterization of activated carbon
The surface area of the activated carbon was determined using a BELSORP-Mini machine. Fourier transform infrared (FTIR-KBR) with the obtained spectra between 450 and 4000 cm −1 was used to determine the surface properties and functional groups. Surface morphology and element composition analyses were performed using a field emission scanning electron microscope (FESEM), and the total ash content of the adsorbent was determined in accordance with ASTM standard D2866 -11 [19].

Characterization results
3.1.1. BET surface area and pore characteristics. Table 1 shows the BET surface area and characteristics for all the samples. Different gas treatments showed varying surface areas. The highest surface area of the non-functionalized activated carbon was obtained from the absence of any gas treatment at 882.71 m 2 /g, whereas the lowest surface area from steam treatment was 842.25 m 2 /g.  These bonds were possibly ascribed to the cerebrin structure of sea mango. The N-H bonds from the primary amine group only existed in the DES-functionalized activated carbon because the presence of choline chloride in the activated carbon contributed to the N-H bond. Figure 2 shows the suggested attachment of DES on the surface of the activated carbon. The elemental N which is believed to assist in CO2 capture into the substrate was successfully attached on the surface of the activated carbon as per illustrated in figure 3 [21]. The C-C stretch from the aromatic group was present in the raw and non-functionalized activated carbon; however, this bond was absent in the DES-functionalized activated carbon because the reaction of H from DES with C from cerebrin promoted a new band, which is a C-H bond from the alkane group at peaks 1477.89 cm −1 to 1478.35 cm −1 . The C-O stretch existed in the non-functionalized activated carbon at peaks 1055.02 cm −1 to 1150.09 cm −1 . However, this bond was absent in the DES-functionalized activated carbon. The C-N stretch bond from the aliphatic amine group existed only in the DES-functionalized activated carbon because it is attributed to the choline chloride of DES.
The C-H bond from the alkane group was present at peaks 955.94 cm −1 to 956.58 cm −1 for the DES-functionalized activated carbon but was absent in the raw and non-functionalized activated carbon. The C-H bond was likely attributed to glycerol in DES. The C-Cl stretch from the alkyl halide group was present only in the DES-functionalized activated carbon because of the presence of Cl element in the choline chloride. As such, the functionalization process added several functional groups to the surface of the activated carbon. The presence of DES increases the number of active sites on the surface and enhances the CO2 adsorption capacity [22].    Figure 4(a) shows the physical structure of a raw sample. Evidently, the pore volume was very limited and not well developed. After carbonization and impregnation with H3PO4 were conducted, a honeycomb-like pore structure with a high pore volume was clearly observed, as confirmed by the BET results [figures 4(b) to 4(e)]. The pore structures were developed during physiochemical activation with the help of H3PO4, which assists the development of pores [23]. The well-developed pores resulted in the large surface area and pore structure, which allowed a good surface for CO2 adsorption. These images also confirmed that no chemical was attached to the surface of the activated carbon. Thus, CO2 adsorption occurred between the activated carbon particle surface and CO2 molecules, and this process is known as physisorption.
After impregnation with DES was completed, few pores were found clogged. Consequently, the pore volume decreased. CO2 adsorption occurred primarily through chemisorption as the DES on the surface of activated carbon assisted CO2 adsorption; thus, CO2 adsorption in the activated carbon was enhanced [22].

Energy-dispersive X-Ray (EDX) analysis.
EDX is an X-ray technique used to identify the elemental composition of materials. Elemental analysis revealed that the raw, non-functionalized, and DES-functionalized activated carbon contained C, O, P, N, and Cl (table 3). Approximately 58.00% to 88.00% carbon content was found in all the samples related to the original content in sea mango (Cerbera odollam). After the raw material was activated, the composition of C increased significantly from 66.00% to 88.00%. This finding showed that the activation process helped increase C composition because heat supply during activation initiates thermal degradation; thus, volatile matter is removed and a stable carbon element is retained [24]. P was present in the non-functionalized activated carbon and DES-functionalized activated carbon, and this finding indicated a reaction with H3PO4 during the impregnation process. N and Cl were also detected in the DES-functionalized activated carbon possibly because of the presence of choline chloride of DES. This finding confirmed that N was successfully attached to the surface of the activated carbon; as a result, more active sites become available for CO2 adsorption, and CO2 capture is enhanced [22].

Breakthrough time of CO2 adsorption
The CO2 adsorption test was conducted to determine the breakthrough time of each sample. Figure 5 illustrates the breakthrough time curve for the non-functionalized sample by using different types of flowing gas treatment. In figure 5(a) and table 4, the highest CO2 adsorption capacity and the longest breakthrough time of the non-functionalized activated carbon were observed in the AC-AAG sample at 9.585 mgCO2/gsol, followed by AC-S at 9.485mgCO2/gsol, AC-N2 at 9.286 mgCO2/gsol, and AC-CO2 at 8.412 mgCO2/gsol. The highest adsorption capacity was detected in AC-AAG because it exhibited the highest surface area, which allows more CO2 molecules to bind to the active sites of the activated carbon [5].  For the DES-functionalized activated carbon [ figure 5(b) and table 4], the highest CO2 adsorption capacity was found in AC-CO2-D at 9.851 mgCO2/gsol, followed by AC-AAG-D at 9.685 mgCO2/gsol, AC-S-D at 9.636 mgCO2/gsol, and AC-N2-D at 9.536 mgCO2/gsol. The highest CO2 adsorption capacity was observed in AC-CO2-D because of its surface area and available elemental N on the surface, which significantly accelerated CO2 adsorption [21]. Hence, the DES-functionalized activated carbon demonstrated a more efficient performance than the non-functionalized activated carbon did.

Conclusion
The DES-functionalized activated carbon yielded a higher CO2 adsorption capacity than the nonfunctionalized activated carbon did because the elemental N on the surface of the activated carbon increased the number of active sites for CO2 capture. Although the surface area of the DESfunctionalized activated carbon was reduced to almost half of the non-functionalized activated carbon, the adsorption capacity of the DES-functionalized activated carbon was improved. Therefore, the DES-functionalized activated carbon underwent chemisorption in its active site. Elemental N from DES played a vital role in CO2 adsorption.