Abstract

A new CO2 adsorbent is produced from waste oil fly ash (OFA). Ammonium hydroxide solution is used to convert OFA to activated carbon. Then, the product is used for the adsorption of CO2 from a nitrogen/carbon dioxide (N2/CO2) gas mixture. The OFA samples are characterized by several techniques. Chemical treatment of OFA considerably changed its surface morphology. In particular, its surface area, as determined by BET measurements, increased from 59 to 318 m2/g. The amine-functionalized ash had a monolayer adsorption capacity of 74.51 mg/g and was obtained at relative pressure, . A kinetics study showed that the CO2 adsorption capacity of OFA increased with increasing CO2 flow rates and concentrations and decreasing the relative humidity. Unlike physical adsorption, the chemisorption process resulted in increased adsorption capacity with increasing temperatures over the range 0–40°C. We also found that the adsorption process was endothermic (80–173 kJ/mol). The isotherm data for the adsorption process were fitted using different models. The saturation capacity determined from the Sips model, which corresponds to the sum of the saturation capacities of all of the adsorbed layers, was 540.3 mg/g of ash.

1. Introduction

The emission of CO2 from the combustion of fossil fuels and other sources has recently increased the CO2 concentration in the atmosphere to an alarming level of 400 ppm [1]. Human activities result in the emission of approximately 29 billion tons of CO2 into the atmosphere per year, where it acts as a heat-trapping greenhouse gas. Among all greenhouse gases, CO2 is potentially the highest contributor to the climate change [2]. Stabilizing the CO2 content of the atmosphere can potentially be achieved by the capture and sequestration of CO2 from its primary emission sources. CO2 can be recovered using processes such as membrane separation, liquid solvent absorption using amines, and pressure/temperature swing adsorption or cryogenic techniques [3]. However, the amine-based processes that are currently used in chemical plants to separate CO2 require large amounts of water and are energy intensive [1]. Additionally, these processes are associated with the generation of byproducts, corrosion of processing equipment, and a high energy of regeneration [4, 5].

Solid adsorbents also have potential to be used in separating CO2 from gas stream, because they require relatively little energy and can be applied easily over a wide range of temperatures and pressures [59]. However, the effectiveness of this technique depends on the development of durable and easily regenerable adsorbents with high selectivity and adsorption capacities [10]. Such materials include activated carbons, carbo-aluminosilicates, zeolites, fly ash, and metal oxides [11]. Promising results are obtained using zeolites to separate CO2 from gas streams. At low temperatures (20–50°C), these aluminosilicate-based materials are known to exhibit a high adsorption capacity because of their large internal pore volumes, structural features, molecular pore sizes, and wide range of chemical composition [12]. However, the adsorption capacities of these materials decrease rapidly as the temperature increases [13, 14]. Moreover, the adsorption ability of aluminosilicates is affected by the presence of water due to hydrophilic nature of these materials [15]. Unlike zeolites, activated carbon does not require moisture removal and has a large surface area and an easily modifiable pore structure. However, the use of carbon-based materials as an adsorbent for CO2 separation is limited as a result of the high sensitivity of these materials to the temperatures that are associated with power plant flue gases [1].

Fly ash (FA) is a byproduct of the combustion of fuel oil in power plants that is normally collected using cyclones (mechanical devices) or electrostatic precipitators. It is estimated that worldwide generation of fly ash is 750 million tones where China is the largest producer with amount of 580 million tons in 2015 [16]. FA can be used in a range of applications, including water pollution control [17], agriculture [18], and metal recovery [19, 20]. However, FA is a major waste material and requires proper disposal. Worldwide, approximately 25% of the fly ash that is produced annually is used. In the United States and China, large amounts of fly ash are produced, at utilization levels of 32% and 40%, respectively [21, 22]. Thus, fly ash is an abundant and economical adsorbent precursor that can be used to treat flue gases as well as recover metals from wastewater [23]. Several researchers have used fly ash for these purposes. For instance, Mercedes Maroto-Valer et al. [24] used amine-enriched fly ash to separate CO2 from a gas stream. Panday et al. [25] demonstrated the suitability of using ash from a thermal power plant for the removal of copper from aqueous solutions. The copper adsorption capacity of the solid adsorbent (ash) was dependent on the concentration, temperature, and pH of the solution.

Recently, our group developed a new activated OFA using various oxidizing agents such as sulfuric and nitric acids and the resulting products were used for CO2 capture from CO2/N2 gas stream [26]. The surface area and adsorption capacity were enhanced in comparison with untreated OFA. In the present study, OFA is subjected to an alkaline treatment using ammonium hydroxide solution in an attempt to improve the adsorption capacity of CO2 on the surface through chemisorption in addition to physisorption. The performances of the amine-treated and nontreated OFA are compared. The previous work discussed the synthesis of the adsorbent and in this study the focus is on the kinetics and thermodynamics of CO2 sorption onto the alkaline treated OFA. Further, the CO2 sorption on alkaline-treated OFA is modeled.

2. Methods

2.1. Materials

The OFA used in this study was produced by local electric power generation plants. Ammonium hydroxide solution (30% NH3 in water, Sigma-Aldrich) was used for the chemical treatment and had a density of 0.9 g/mL at 25°C and a purity of ≥99.99%. A portable CO2 meter (from TESTO CO2) and GC-MS (Agilent Technologies) were used to measure the CO2 outlet concentrations and relative humidity was measured by a relative humidity meter at the bed inlet and outlet. Deionized water and soap were used to wash all glassware to remove any adhered impurities.

2.2. Activation of OFA

Ammonium hydroxide solution was used for the activation of OFA. A representative sample of 100 g of the ash was mixed thoroughly with 300 mL of ammonium hydroxide solution at room temperature and then transferred to a 500 mL flask and refluxed at 120°C for 24 h using a temperature ramp of 1°C/min. The mixture was cooled, after which 150 mL of ammonium hydroxide was added and mixed for another 24 h. The resultant mixture was filtered and the filter cake was dried in the oven at 105°C for 4 h and then left to cool at room temperature. The presence of amine functional groups on the OFA surface (as a result of ammonium hydroxide treatment) was confirmed by FTIR analysis of the final product.

2.3. Characterization and Analysis of OFA and Activated OFA

The morphology of OFA was characterized by scanning electron microscopy (SEM) on a JEOL JSM 6400 scanning electron microscope. Prior to analysis at different magnifications, the samples were coated with carbon. The surface area was determined by nitrogen adsorption at 77 K on a Micrometrics 2020 instrument. Prior to analysis, the samples were degassed under vacuum and dried at 150°C for 6 h. A PerkinElmer FTIR spectrophotometer was used to analyze the functional groups of the samples.

2.4. Adsorption Studies
2.4.1. Adsorption Isotherm Study

First the adsorption isotherm of CO2 was measured at different initial pressure values using a magnetic suspension balance (Robotherm). A sample of approximately 0.35 g of OFA was degassed at 100°C for 5 h and subjected to a buoyancy measurement using helium at 25°C. The adsorption isotherm was obtained by measuring the equilibrium weight of the sample at 25°C at various equilibrium pressures of pure carbon dioxide (1–40 bar).

2.4.2. Adsorption Kinetics Study

Next, we measured kinetics parameters including the initial CO2 concentration (363 and 795 ppm), flow rate (0.3, 0.7, and 1.0 mL/min), humidity (20% and 60% RH), and bed temperature (−20 to 40°C) using a jacketed quartz adsorption column (18 cm in length and 1.0 cm in diameter), whereby a glycol/water mixture was circulated through the jacket to maintain isothermal conditions using a PID temperature controller (Figure 1). A representative sample of 1.5 g of activated OFA was placed in the adsorption column. Nitrogen was then introduced into the bed and the concentration of CO2 was measured to ensure that no CO2 was initially present in the ash sample. Subsequently, a CO2/N2 mixture was introduced into the column at different flow rates, concentrations, temperatures, and relative humidities. At the same time, the concentration of CO2 exiting the column was measured at different time intervals. The relative humidity was measured at the inlet and outlet of the column. N2 gas was used in the desorption experiments following the same procedure as that used for the adsorption experiments.

3. Results and Discussion

3.1. Characterization of Activated Fly Ash

The SEM images of the untreated and treated OFA samples are shown in Figure 2. As observed, the ash samples mostly consisted of spherical particles. The untreated ash contained large and small pores, with an average pore diameter of 2-3 μm. The particle size ranged from 10 to 100 μm [27]. In contrast, the morphology of the activated OFA exhibited a well-defined porous structure in which the pores were more open than those present in the original OFA.

The FTIR spectrum of the untreated ash is given in Figure 3. The untreated ash exhibited different peaks at 1123, 1623, 2371, 3307, and 3772 cm−1. These broad peaks were associated with C=C, C=O, C=C, H–C=, and O–H functional groups, respectively. The chemically treated ash displayed a broad peak at 3402 cm−1 that indicated the presence of –NH2–, while the peak at 3186 cm−1 was attributed to –NH2 symmetric and –NH2 asymmetric stretching vibrations [28]. However, after chemical activation, the intensities of the untreated ash peaks at 3772 cm−1 and 3206 cm−1 were reduced to 3402 cm−1 and 3186 cm−1, respectively. At a high alkaline pH, the OH groups attack the surface of OFA during the activation process. Moreover, during the activation process in hot solution, chemical reactions involving carbon oxidation and hydroxide reduction are expected to occur between OFA and ammonium hydroxide. This reaction results in the expansion of the interlayers of adjacent hexagonal network planes in the OFA structure.

Figure 4 shows the N2 adsorption-desorption isotherms measured at 77 K of the untreated and chemically treated ash samples. The BET surface area increased from 59 m2/g for the untreated ash to 318 m2/g for the OFA after chemical activation. The pore volume also increased substantially from 0.0368 to 0.6786 cm3/g following chemical activation (Table 1). Reinik et al. subjected OFA to a hydrothermal treatment in the presence of 5 M NaOH. The resulting OFA displayed a high surface area of 67.9 m2/g [29]. In a recent study, Mofarrah et al. treated OFA with various oxidizing agents, that is, nitric, hydrochloric, and phosphoric acids, followed by physical activation at 800°C. The researchers obtained activated OFA with a surface area of 123.19 m2/g and a total pore volume of 0.2215 cm3/g [30].

Table 2 shows the elemental analysis of OFA before and after chemical treatment. It can be seen that the fly ash before treatment was composed primarily of carbon with some sulfur and traces of heavy metals [31, 32]. Unlike the effects of acid treatment, the alkaline treatment did not result in noticeable changes in the composition of the fly ash. The sodium in the fly ash either leached out upon treatment or was not detectable by EDX.

3.2. Adsorption Isotherm of CO2 Using Activated OFA

The CO2 adsorption-desorption isotherm of activated OFA obtained at different pressures using the Rubotherm instruments is shown in Figure 5. The amount of CO2 that was adsorbed per gram of OFA was higher than that desorbed at the same equilibrium pressure. A multilayer adsorption process occurred as the equilibrium pressure increased. The difference between the adsorption and desorption curves was related to the amount of CO2 that was chemisorbed onto the OFA surface. Initially, at low equilibrium pressures, a high amount of CO2 was chemisorbed, reaching a maximum value of 17.8 mg/g at 16 bar, followed by a decrease in the uptake of CO2 as the equilibrium pressure was further increased. Increasing the sample pressure increased the concentration of CO2 molecules on the active sites that resulted in more CO2–surface interactions. Further increases in the pressure reduced the amount of CO2 chemisorbed onto the surface, due to consumption of the reactive amine sites and the attainment of the saturation adsorption level in the OFA pores (Figure 6).

The Langmuir, Freundlich, and BET isotherm models were examined to describe the adsorption process and the CO2 adsorption capacity of treated OFA was estimated accordingly.

The Langmuir model describes monolayer coverage on a solid surface as follows:where (mmol/g) is the maximum amount of CO2 that is adsorbed per unit mass for a complete monolayer, is the equilibrium pressure of CO2 (bar), and (L/mmol) is the Langmuir constant that is related to the affinity of the binding sites of the adsorbent.

The Freundlich model accounts for surface heterogeneity in the energy and is represented by (4):where both and are constants.

The Sips model is a combination of the Langmuir and Freundlich models. The Sips model incorporates both the effect of surface heterogeneity and the saturation capacity of the adsorbent. At low equilibrium concentrations of the adsorbate, the Sips model reduces to the Freundlich model, whereas at high concentrations the amount of adsorbed solutes reaches a plateau at a saturation value,  (mmol/g) [33]. The Sips model is described by the following expression:In the BET model, the formation of multiple layers on the surface of the adsorbent is assumed, thereby achieving a multilayer coverage on the material surface:where , , and represent the saturation capacity, constant account for difference in adsorption energies between the first layer and higher layers, and saturation pressure for CO2 at the studied system temperature. The corresponding isotherm values were fitted using MATLAB (V. 8.3) and the results are presented in Table 3. Among all the studied isotherm models, the BET model best fits the adsorption data. The model generated a high regression coefficient of 0.990 and the lowest sum of square errors of 13.44. The saturation adsorption capacity determined from the BET model, which corresponds to a monolayer of adsorption in pressure range , was 74.51 mg/g of ash. The saturation capacity determined from the Sips model which corresponds to the sum of the saturation capacities of all of the adsorbed layers was 540.3 mg/g of ash.

3.3. Analysis of the Adsorption Kinetics

The rate of adsorption was determined by passing CO2 through the adsorption column at fixed initial concentrations and relative humidity and monitoring the concentration of CO2 exiting the column as a function of time. Kinetics parameters, namely, the bed temperature, gas flow rate, and concentration and humidity, were investigated. Desorption experiments were performed in the same manner as the adsorption measurements, except that pure and dry N2 gas was used.

Figure 7 shows the breakthrough curves obtained using 1.5 g of treated OFA, an initial CO2 concentration of 350 mg/L, and a CO2/N2 flow rate of 0.3 L/min and at 20% relative humidity (RH), 1 atm, and 22°C. The areas above the adsorption and desorption curves correspond to the adsorbed and desorbed amounts of CO2, respectively. The area between the two curves corresponds to the amount (in mg) of CO2 that remained on the OFA surface. The trapezoidal rule was used to estimate the amounts of adsorbed (0.238 mg/g) and desorbed (0.115 mg/g) CO2 at atmospheric pressure. The difference of 0.123 mg/g between the aforementioned values corresponded to the amount of CO2 that was chemisorbed onto the ash bed.

3.3.1. Effect of Initial CO2 Concentration

Further results on the effect of the initial CO2 concentration are shown in Figures 8 and 9. The OFA adsorption capacity increased with increasing CO2 concentrations [34, 35] because of the increase in driving force between the bulk concentration and the concentration at the particle surface, thereby promoting mass transfer. However, when the ash surface was saturated, further increases in the CO2 concentration did not further improve the adsorption capacity. This activated OFA reached its saturation capacity at higher concentrations quickly. Thus, the saturation time decreased as the adsorbate concentration increased [36, 37].

3.3.2. Effect of Flow Rate

The effect of the gas flow rate was studied at two different CO2 concentrations, 795 ppm and 365 ppm, at both low and high humidities. The effect of the CO2 flow rate was studied using flow rates of 0.3, 0.7, and 1.5 mL/min. Figure 10 shows that increasing the CO2 flow rate reduced the time required to reach the maximum adsorption capacity. Hence, the residence time of the gas in the ash bed decreased [34]. Increasing the gas flow rate increased the diffusion rate of the CO2 molecules to the particle surface. In addition, the rate of mass transfer in the gas film was likely enhanced because of the reduction in the external mass transfer resistance and the saturation time [38, 39]. It has also been reported [39, 40] that the application of CO2 at high pressures increases the strength of the interaction between amine and CO2 molecules, which favors the chemical reaction and increases the extent of chemisorption.

3.3.3. Effect of Humidity

The effect of relative humidity (10% and 80%) on the adsorption of CO2 onto treated fly ash is shown in Figure 11. A decrease in the adsorption capacity at the higher relative humidity was observed. This result can be explained by consideration of the capillary condensation of water vapor in the pores of the ash; at 80% RH, a large portion of the active sites on the surface of these pores are blocked [41, 42]. Increasing the humidity decreases the partial pressure of CO2 in the gas phase and leads to water vapor competing with CO2 molecules to reach the surface. When water condensation dominates over evaporation, there is competition between these two species for adsorption onto the active OFA sites. This was confirmed by measuring the RH at the column’s exit; a 2% decrease in relative humidity was observed. Increasing the adsorption capacity could be achieved by increasing the temperature while reducing the humidity, thus opening up the mesopores [43].

3.3.4. Effect of Temperature

Next, we investigated the effects of bed temperatures of −20, −10, 0, 10, 20, 30, and 40°C on the rate of CO2 adsorption onto treated OFA. We found that the adsorption rate of CO2 increased up to a break-even point as the bed and inlet gas temperatures increased (Figure 12). This demonstrates the endergonic nature of the interaction between the gas molecules and the ash surface at which both physical and chemical adsorption occurred. Moreover, there was a direct link between increases in the temperature and the pressure drop across the bed [44].

Equations (1) and (2) were used to determine thermodynamic parameters such as the heat of adsorption, standard entropy, and standard free energy of the adsorption process [4547]:where is the heat of adsorption, is the standard entropy, is the standard Gibbs free energy, and and are the adsorbed and solution equilibrium adsorption values, respectively. is the gas constant (8.314 J/mol K).

The heat of adsorption and the standard entropy were obtained by plotting versus 1/ at both low and high humidity (Figure 13). and were calculated from the slope and the intercept of the plot, respectively [27, 4550].

As the temperature increased, both and the adsorption capacity increased, which could be attributed to the increase in the number of active surface sites that were available for adsorption. Additionally, the motion of CO2 molecules on the ash surface increased owing to increases in the temperature [27]. The calculated values of , , and are listed in Table 4. The positive value of at a fixed RH indicated that the adsorption process was endothermic [45, 50]. The positive value of showed the affinity of the chemically treated fly ash for CO2 (adsorption) and the increasing randomness during the adsorption process [51]. The negative values of at −10 and −20°C indicate the spontaneous nature of the adsorption of CO2 onto the fly ash sample [49]. Positive values were observed at 0, 10, 20, 30, and 40°C, indicating that spontaneity was not favored at high temperatures where the thermal energy exceeded the adsorption energy. A similar trend was observed for the adsorption/removal of Congo Red from water using activated carbon that was prepared from coir pith [50]. Moreover, increased with increasing temperatures, further indicating that the reaction was endothermic [52].

4. Conclusions

We suggest that waste OFA, which is primarily composed of unburned carbon, can be used as a low-cost adsorbent for the removal of CO2 from a gas stream. Promising results were obtained using chemically treated waste OFA for CO2 adsorption. An ammonium hydroxide solution was used to chemically modify fly ash to improve its CO2 adsorption capacity. The BET surface area of OFA increased from 59 m2/g to 318 m2/g after chemical treatment. Studies of the underlying kinetics showed that the adsorption capacity increases with flow rate, concentration, and temperature. Humidity also strongly influences CO2 adsorption. Thermodynamic studies revealed the heat of adsorption to be 80 and 178 kJ/mol at low (20%) and high (80%) relative humidity, respectively, indicating that the process was endothermic. This indicated that the adsorption of CO2 in alkaline treated OFA is sensitive to humidity and suggests the competitiveness between water vapor and CO2 adsorption on the surface. A variety of isotherm models were studied to obtain the saturation capacity of the treated OFA and it was found that 540.3 mg/g can be predicted using Sips model, and 652.9 mg/g is achieved by Langmuir model with regression coefficients 0.982 and 0.968, respectively. The activated carbonaceous material presented herein is a promising and economically feasible CO2 adsorbent because of its low cost and high sorption capacity.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgments

The authors would like to acknowledge financial support from King Abdul Aziz City for Science and Technology (KACST) through the Science and Technology Unit at King Fahd University of Petroleum & Minerals (KFUPM), under Project no. 11-ENV1645-04. The Gas Processing Center at Qatar University is also acknowledged for contributing to the analysis and characterization of the studied material.