Investigation of methane adsorption on molecular sieve zeolite (from natural materials)

1. Introduction

Among the different technologies used for the separation and capturing of carbon dioxide and methane gases, the adsorption process is increasingly competitive because of low maintenance, cost-effectiveness, low energy requirement, and ease of use. (Pires et al., 2008). Various types of adsorbents such as alumina (Delgado et al., 2006), carbons (Llano-Restrepo, 2010; Lee and Park, 2015), metal-organic frameworks (MOFs) (Alaie et al., 2015; Hesas et al., 2015), and molecular sieves (Chen et al., 2014) are used for the removal of CH₄ and CO₂ from the gas stream.

Among the adsorbents, molecular sieves and zeolites are widely considered due to high stability, performance, and low cost. (Lee et al., 2002; Alshameri et al., 2014; Esfandian, 2015; Esfandian et al., 2016b; Opanasenko et al., 2016; Esfandian et al., 2017; Garshasbi et al., 2017). In molecular sieves, the sodium form of the type X crystal (called 13X) is widely used for gas adsorption processes due to high crystallinity and porosity (Bandarchian and Anbia, 2015; Abate et al., 2016; Bacsik et al., 2016). Generally, zeolites are synthesized from sodium aluminosilicate gel formed from various silica and alumina sources using the hydrothermal method. However, the preparation of synthetic zeolites from the chemical source of silica and alumina is expensive. In order to reduce costs, numerous researches are seeking cheap raw materials such as paper sludge (Wajima et al., 2006; Purnomo et al., 2012; Ansari et al., 2014), slag of lithium (Chen et al., 2012b), high silicon fly ash (Kazemian et al., 2010), shale ash of oil (Machado and Miotto, 2005), coal fly ash (Tanaka and Fujii, 2009) bentonite (Ma et al., 2014a), kaolinite and waste porcelain (Ma et al., 2014a; Wang et al., 2014). Among the mentioned sources, Feldspars, kaolin, and bentonite are used for the synthesis of 13X due to the lower costs, availability, and some special properties (Blatt et al., 2006; Holmes et al., 2011; Chen et al., 2014; Ma et al., 2014a). Several studies reported the synthesis of the 13X from natural sources but the reported adsorption capacities were not considerable.  Ma et al. reported the synthesis of 13X from low-grade natural kaolin via alkali fusion followed by a hydrothermal process. The effect of different parameters, such as NaOH addition amount and crystallization time, on the crystallinity of products, was systematically investigated (Ma et al., 2014b). In other papers, 13X was synthesized using bentonite as the raw material by alkaline fusion followed by a hydrothermal process and used for CO₂ capturing. The synthesized 13X performance was also compared to the commercial ones; the synthesized sample showed lower adsorption capacity (Chen et al., 2014; Ma et al., 2014a).

In this paper, molecular sieve 13X was synthesized by the hydrothermal method using some natural materials such as kaolin, bentonite, and feldspar to adsorb CH4 over a temperature range. The effects of various operational parameters including molar ratio, temperature, time of reaction on the crystallization, and the purity of the synthesized product have been studied. Furthermore, the properties of the synthesized samples, such as crystal morphology, framework structure, and pore structure were characterized by SEM, FT-IR, XRD, TGA, and BET methods. The experimental data of CH₄ adsorption were also fitted to some isotherm models including Langmuir, Sips, Toth, BET, and Freundlich isotherm models to determine the physicochemical properties, such as the adsorption capacity. Finally, the adsorption capacities of synthesized samples were compared with the other adsorbents.

2. Materials and methods

2.1. Preparation of adsorbent

Due to the high content of SiO₂,  the natural materials (kaolin, bentonite, and feldspar) should be treated by the acidizing method to achieve 13X. Table 1 shows the main properties of raw and modified natural materials. The obtained mixture gel was stirred for 120 h with magnetic stirring at room temperature and then a homogenous solution was transferred into a Teflon-Lined autoclave for crystallization at the temperature of 65 ± 5 °C for 72h. The products were filtered, washed several times with distilled water (until pH of supernatant reached 9), and dried in the oven at 80 °C for 12 h.

Table 1. Natural materials properties

2.2. Characterization Methods

The X-ray Fluorescence (XRF) was used for the characterization of the chemical composition of kaolin, bentonite, and feldspar by a Wavelength X-ray dispersive Fluorescence Spectrometer (Model Bruker S4). The X-ray diffraction (XRD) analysis was employed to characterize the structure of the samples with a GBC MMA diffractometer (CuKa) (model Philips PW1140/90).

The bonds of the natural kaolin, bentonite, feldspar, and molecular sieve 13X were obtained using the FT-IR spectrometer (Perkin-Elmer). The scanning electron microscopy (SEM) was used to investigate the morphological features of the natural and synthesized zeolites. TGA (Thermogravimetric Analysis) was used to check the thermal behavior of the products. Brunauer-Emmett-Teller (BET) surface area analysis was used to calculate pore size distribution, pore diameter, and specific surface area at liquid nitrogen temperature (77 K).

2.3. Experimental Design by Taguchi

Experimental design by the Taguchi method is a well-known tool for the optimization of the multifactor process (Garshasbi et al., 2017). The experimental design of these experiments and the response were achieved using the option of Automatic Design in Qualitek-4 software featuring an L9 orthogonal array with 9 trails. The experimental conditions for the synthesis of the different types of 13X are presented in Table 2.

Table 2. DOE for synthesized molecular sieve according to Taguchi L9 array

2.4. Isotherm Models

The adsorption isotherms are the fundamental requirements for optimizing adsorption systems (Colina et al., 2001). The experimentaldata of CH₄ adsorptiononto 13X samples were fitted to some standard isotherm models to examine the model's constants adsorption isotherms.

2.5. Langmuir isotherm

The Langmuir isotherm model is based on the four assumptions: (a) a fixed number of sites for the adsorption, (b) identical sites, (c) monolayer adsorption, and (d) no interaction between adsorbate molecules. The Langmuir isotherm equation is defined as the following equation (Ladshaw et al., 2015):

Where b and q_mi are the equilibrium constant and monolayer maximum adsorption capacity, respectively.

2.5.1. Sips Equation

Sips isotherm model is based on localized adsorption without considering adsorbate-adsorbate interactions. The Sips isotherm was formed by a combination of the Langmuir and Freundlich expressions. At low concentrations, Sips convert to Freundlich isotherm whereas, at high concentrations, it behaves as the Langmuir isotherm. The adsorption isotherm model is presented as the following equation (Tzabar and ter Brake, 2016):

Where P is the pressure, q_mi (mol/Kg) is the maximum adsorption capacity, b is the sorption affinity, and n as a dimensionless parameter is related to the heterogeneity of the adsorbate-adsorbent process. Q and b₀ are the isosteric heat of sorption at half loading and the constant of adsorption at the temperature  respectively. The dependency of heterogeneity parameter (n) to the temperature is described by  and  parameters.

2.5.2. Toth Isotherm

The Toth isotherm is a semi-empirical expression obtained by extending the Langmuir equation and considering the heterogeneity of adsorbent surfaces (Toth, 2002). The three parameters Toth isotherm is based on a quasi-Gaussian energy distribution which can be introduced as the most useful isotherm in describing the heterogeneous adsorption systems of gases at a wide range of pressure. Toth isotherm model is often given in the form of the following equation:

Where q_mi (mol/Kg) is the maximum adsorption capacity, b (Mpa-1) and n are Toth isotherm constant. It is clear that when 𝑛 = 1 (identical sites), this equation reduces to Langmuir isotherm equation (Tóth, 1995; Toth, 2002).

2.5.3. BET isotherm

For multilayer adsorption, BET adsorption isotherm is one of the most successful isotherm models used to express the adsorption process generated by Brunauer et al. (Brunauer et al., 1938), which assumes the interaction between the solid and the adsorbate (gas) is much larger than that between neighboring molecules. The BET equation is given by Eq.6 (Chen et al., 2017):

Where ,  and b are the maximum loading of gas onto the surface of molecular sieves, the saturation pressure of CH₄ at experimental temperature, and the equation constant that is exponentially related to the energy of monolayer adsorption respectively.

2.5.4. Freundlich isotherm

Freundlich adsorption isotherm is an empirical equation based on the heterogeneity of the adsorbent surface. The Freundlich isotherm is represented by (Freundlich, 1906):

Where n and k are the constants of the Freundlich equation in relationship with the heterogeneity of the adsorption surface and capacity, respectively.

2.6. Thermodynamic Studies

The determination of the adsorption mechanism is very necessary for an adsorption study. Thermodynamic parameters such as Gibbs free energy change (ΔG°), enthalpy change (ΔH°) and entropy change (ΔS°) can be explained using the adsorption mechanism. Thermodynamic parameters were estimated through the following equations (Zhou et al., 2013; Esfandian et al., 2016a):

Where  ( ), R ( ),   ( ) and T are the standard Gibbs free energy change, gas constant, the Langmuir equilibrium constant,  temperature respectively;  ) and  ( ) are defined asa change in standard enthalpy and entropy, respectively.

2.7.  Measurement of gas Adsorption

The volumetric analysis was used to investigate the capacity of molecular sieve 13X for CH₄ sorption (Fig.1). Due to its low cost, simplicity, and easy assemblage, this quantitative analytical method is extensively adopted for the gas uptake measurements. (Colina et al., 2001; Holmes et al., 2011). The CH₄ adsorption processes were explained by details in previous work (Garshasbi et al., 2017).

Figure1. Volumetric set up for gas adsorption test.

3. Results and discussion

3.1. Taguchi Design of Experiment analysis

In these experiments, an L9 array design was carried out to analyze the effect of four variables of Al₂O₃/ SiO₂, Na₂O/Al₂O₃, Time (h), and Temperature (⁰C) for the synthesis of the 13X using the Taguchi method. The relative crystallinity was selected as the test response. Fig. 2 shows the best levels for each factor obtained from each experiment (Taguchi L9 array) for the synthesis of 13X. After synthesis of the 13X zeolite powder, their properties were characterized. The synthesized powders crystallinities were considered as the desired response. Using ANOVA parameters where have significant effects on the synthesized 13X zeolite powder crystallinity can be selected. The main objective of ANOVA analysis is further investigation of the significance of the selected parameters in terms of their impact(s) on the concerned response(s), i. e. the 13X zeolite content of produced powders here. As observed from Table 3, the most affective synthesis parameter of the 13X zeolite content is the SiO₂/Al₂O₃ ratio (with minimum P-value or maximum F-value). The second and third parameters, which great impact on the 13X zeolite formation, are the Na₂O/Al₂O₃ ratio and the temperature of crystallization respectively. Finally, the duration of crystallization was found as the lowest significant effect on the synthesized zeolite crystallinity.

Figure 2. Best levels for selected parameters.

Table 3.Analysis of variance (ANOVA) of carried out experiments in terms of the synthesized 13X zeolite.

3.2.        Characterization of Synthesis Molecular Sieve 13X

Fig.3 indicates the XRD pattern of natural kaolin, modified kaolin, bentonite, modified bentonite, feldspar, modified feldspar, and commercial molecular sieve 13X. The modification was done by giving heat to natural materials at 900 ˚C for 2 h.  XRD pattern shows a significant change between modified and untreated natural materials. The characteristic peaks of synthesized products attributed to 13X are observable at 2θ values of 6.12˚, 10.00 ˚, 23.31 ˚, 26.65 ˚, 33.59 ˚, and 37.34 ˚.

The FTIR of samples is shown in Fig. 4. The results indicated that the broadband of natural materials is in the range of about 920 cm^-1 to about 670 cm^-1 and is related to Al-O bonds in Al₂O₃. By the reaction between NaOH and SiO₂ and Al₂O_3, aluminosilicates were produced. The bending vibration bands in the IR spectrum were replaced by a single band around 1000 cm^-1, characteristic of Si–O–Al bonds (Cavenati et al., 2004).

The morphology of the natural materials and synthesized 13X are presented in Fig.5. As shown, the 13X produced from natural kaolin, bentonite, and feldspar showed that the formation of spherical micro-sized particles with an average size of 5 μm.

Figure 6 shows the TGA profile of the 13X samples. According to the results, 13X-K was more resistive to heat.

Figure 3. X-ray diffraction (XRD) pattern of the a) natural kaolin, b) modified kaolin, c) bentonite, d) modified bentonite, e) feldspar, f) modified feldspar and g) commercial molecular sieve 13X.

Figure 4. IR spectra of the natural kaolin, bentonite, feldspar, and molecular sieve 13X.

Figure5. SEM images a) natural kaolin, b) natural bentonite, c) natural feldspar, d) 13X-F,e)13X-B and f) 13X-K.

Figure 6. TGA profile of synthesized molecular sieves.

Figure 7. N2 adsorption-desorption isotherm of synthesis molecular sieves 13X prepared from kaolin (13X- K), prepared from bentonite (13X- B), prepared from feldspar (13X- F) and commercial molecular sieve 13X (13X – C).

Fig.7 shows N₂ adsorption-desorption isotherm of synthesized samples. The specific surface areas and micropore volume were determined by assuming the BET equation from the t-plot method. The BET results of synthesis molecular sieve 13X are presented in Table 4.

Table 4. Molecular sieve 13X textural properties prepared from different natural sources.

3.3. Methane Adsorption

The synthesized 13X samples were used to adsorb the CH₄ at different temperatures (298, 308, 318, and 328 K) and pressures (Fig.8). Results show that the pore structure of molecular sieve 13X was large enough; the steric effect of the adsorbate with the adsorbent structure was ignored. As shown, there was good harmony between the experimental data of CH₄ adsorption in this study with the data found in the literature (Cavenati et al., 2004). The correlation between the CH₄ adsorption vs. pressures at different temperatures (298, 308, 318 and 328 K) for the synthesis of the molecular sieve 13X-K well fitted by following equations (Fig. 9):

The results show that there is a fair correlation of the amount of adsorption on an adsorbent with increased adsorption for lower temperatures.

3.4. Isotherm study

The equilibrium results were modeled and evaluated through five different isotherms and error functions. Based on the maximizing the correlation factor, the comparison of the error function was made, and the best isotherm equations were found. Different isotherm parameters for different types of 13X are shown in Table 5-7. The results indicated that the Sips isotherm provided the best fit to the equilibrium data. There was a good agreement between the predicted theoretical breakthrough curves and the experimental results with Sips> Langmuir > Freundlich > Toth > BET.

Among the two-parameter isotherm models, Langmuir showed the better fitting for all adsorbents. This clearly indicated that CH₄ adsorption took place monolayer onto the homogeneous surface of 13X, which exhibits a uniform distribution of adsorption energy. It was observed that the three-parameter isotherm models (especially Sips) fitted better with adsorption data in comparison to the two-parameter models, illustrating the complexity of the adsorption process. It can be described that Langmuir is the most reliable two-parameter isotherm model used to describe the adsorption of CH₄ on the synthesized adsorbents, while Sips fitted better based on the extra parameter that improved their flexibility and robustness. It should be mentioned that among the three-parameter isotherm models, only the q_max of Sips showed values close to the experimental q_e. (Fig.8). The better fitting of the Langmuir and Sips confirmed the fact that Langmuir is a special case of Sips (Adelodun et al., 2016).

Figure 8.  Experimental data of Methane adsorption on 13X synthesized molecular sieves at different temperatures.

Figure 9. CH4 adsorption behavior on synthesized 13X- K at different temperatures.

A) 298K (B) 308 K, (C) 318 K and (D) 328K.

Table 5. The parameters of CH4 adsorption isotherms at different temperatures on molecular sieve 13X - K

Table 6. The parameters of CH4 adsorption isotherms at different temperatures on molecular sieve 13X – B

Table 7. The parameters of CH4 adsorption isotherms at different temperatures on molecular sieve 13X – F

3.5. Thermodynamic studies

Gibbs free energy ( ), the enthalpy change ( ), and entropy change ( ) for the CH₄ adsorption by 13X were calculated. So, the gas thermodynamic behavior was investigated using the change in free energy ( ), enthalpy ( ) and entropy ( ) (Eq.9-10). Table 8 presents the thermodynamic behavior of CH₄ adsorption. The positive value of  revealed the increased randomness at the solid-solution interface during the gas adsorption. The negative values of  showed the adsorption process was exothermic. Otherwise, the  negative values indicated the spontaneous nature of the adsorption of CH₄ on the adsorbent (Zhang et al., 2010; Chen et al., 2012a).

Table 8. Values Gibbs free energy as a function of the adsorption temperature.

3.6. Comparison of the CH4 adsorption capacities by various sorbents

Table 9 presents the comparison of the adsorption capacity of the synthesized 13-X samples in this study to other literature. As can be seen, in this study, the adsorption capacity of 13X-K was considerably higher than 13X-B and 13X-F. In comparison to the other literature, 13X-K also indicated higher adsorption capacity; however, its adsorption capacity was slightly lower than Acid treated carbon molecular sieve and MOF-235. This comparison and synthesis procedure (low coast (cheap chemical source of silica and alumina) and simple process for the synthesis) indicated that the 13X-K has a high potential to consider as a suitable adsorbent for CH₄ separation in the large scale.

Table 9. Comparison of the CH4 adsorption capacities by different adsorbents at 25 ºC.