Abstract

Dynamic adsorption and desorption of carbon dioxide (CO₂) from binary gas mixtures (CO₂/CH₄ and CO₂/C₃H₈) was carried out on BaX and NaX zeolites in order to study the influence of methane and propane on the CO₂ adsorption capacities and on the preferential adsorption sites of CO₂. The adsorption proprieties of CH₄ and C₃H₈ molecules on the adsorbents were also evaluated. This study enabled to conclude that methane has no affinity with NaX and BaX zeolites in our experimental conditions. On the other hand, whatever the adsorbent studied, the results show favorable adsorption sites to trapping of propane. The adsorption capacities of propane at saturation on NaX and BaX zeolites are 0.035 and 0.075 mmol g^-1 respectively. However, BaX zeolite is the adsorbent which presents more affinity for propane with the strongest propane/zeolite interactions. The study of CO₂ adsorption in presence CH₄ and C₃H₈ on the NaX zeolite indicates a decrease of the CO₂/cation interactions and of adsorbed CO₂ amounts at saturation. The capacities loss are about of 17% and 30% when carbon dioxide adsorption is carried out in binary mixture CO₂/CH₄ and CO₂/C₃H₈ respectively. In addition, a loss of preferential adsorption sites of carbon dioxide between 443 - 483 K is also observable in the presence of CH₄ on NaX zeolite. However on BaX zeolite, the presence of CH₄ and C₃H₈ improves the accumulation of CO₂ in the porosity. The capacities of CO₂ adsorption are improved about of 18% whatever the gas mixture studied. Moreover, contrary to NaX zeolite, a conservation of preferential adsorption sites of CO₂ on BaX zeolite in the temperature range studied (295 - 623 K) is visible. It seems that the presence of methane and propane doesn’t affect the adsorption proprieties of carbon dioxide on BaX zeolite in our experimental conditions.

1. Introduction

According to the latest report of the Intergovernmental Panel on Climate Change (IPCC), carbon dioxide (CO₂) and methane (CH₄) are the two main greenhouse gases (GHGs) responsible of the increase in global temperatures. Indeed, they represent more than 80% of GHG emissions, the majority of which come from fossil fuels ¹. In contrast, carbon dioxide concentrations are 220 times higher than methane concentrations in the atmosphere and have attained their highest level in recent years ¹. CO₂ is thus becoming the main cause of global warming. The urgency to find effective methods and techniques to limit CO₂ emissions into the atmosphere is necessary.

However, technologies are available to capture or separate CO₂ from industrial off-gases in order to limit its emissions ^{2, 3}. Separation techniques such as chemical and physical absorption, cryogenics, membrane technology and physical adsorption are the most common in industry ². In contrast, adsorption techniques, using porous solids as adsorbent, appear to be the most effective for the capture and separation of CO₂ from flue gases ^{2, 4, 5, 6}. Solids with three-dimensional porous structure, such as zeolites, are among the most suitable adsorbents for CO₂ capture or storage ^{7, 8, 9, 10}. On the other hand, faujasite X type zeolites are the most used because of their large micropore volume [10-19].

Indeed, the improvement of such emission reduction and/or carbon dioxide recovery processes requires the search for adsorbents with high capacities of adsorption and high selectivity of CO₂ in the presence of other gases (CH₄, H₂, N₂, H₂O,...) or impurities, low regeneration conditions and good stability in the adsorption/desorption processes. Known the characteristic CO₂ adsorption parameters of the adsorbents is also important in order to improve the selective trapping or separation of carbon dioxide in gas mixtures. Thus, several studies on the separation of carbon dioxide in gas mixtures using X-type zeolites have been conducted ^{20, 21, 22, 23, 24, 25, 26, 27}. Studies on the selectivity of CO₂ in the presence of gases such as CH₄, H₂, O₂ and N₂ have been performed ^{20, 21, 25, 26}. The influence of these gases on the diffusion and capacities of CO₂ adsorption was also evaluated. Other studies on the effect of the presence of water on the capacities of and preferential adsorption sites for CO₂ at low temperatures have also been carried out on faujasite X type zeolites ^{28, 29}. However, all these studies do not show the influence of molecules such as CH₄ and C₃H₈ on the preferential CO₂ adsorption sites at low temperature and on the CO₂ adsorption capacities of X-type zeolites when the carbon dioxide is adsorbed in a binary gas mixture.

Therefore, the objective of this work is to study the influence of the presence of CH₄ and C₃H₈ on the adsorption capacities and on the preferential adsorption sites of CO₂ of NaX and BaX zeolites, during the adsorption of CO₂/CH₄ and CO₂/C₃H₈ gas mixtures. The diffusion of CO₂ into the porosity of the adsorbents as well as the affinity and maximum desorption temperatures of methane and propane molecules will also be evaluated. The fixed bed dynamic adsorption/desorption technique will be the main adsorption technique used in this study.

2. Experimental

The starting material is a NaX zeolite supplied by Axens. The replacement of the sodium cations by barium cations has been obtained by cationic exchange, using barium nitrate Ba(NO₃)₂ from Sigma-Aldrich (purity > 99%). Cationic exchange was carried out at 343 K by stirring 3 g of zeolites powder in 250 mL of aqueous solution containing [Ba²⁺] = 0.02 mol.L^-1 of the metal nitrate for 24 h. The sample was then filtered, washed with ultrapure water (3 x 40 mL) to eliminate the free nitrate ions into the solids. It was then dried in a furnace during 24 h at 373 K. Exchanged BaX zeolite was finally sieved to obtain a particle size between 0.2 and 0.4 mm.

Elemental analysis of M-X zeolites was performed by ICP to quantify the amounts of ion-exchanged cation species.

The thermal stability (mass loss) was performed on an SDT-Q-600 TA instrument with a stream of dry argon (100 mL.min^-1) in a temperature range between 298 and 1073 K, using a ramp of 5 K.min^-1.

Measurements of surface area and pore volume were carried out using a Micromeritics TRISTAR 3000 instrument. About 100-150 mg of the zeolite sample was pretreated during 1 h at 363 K, then 10 h at 623 K. The adsorption/desorption isotherm of nitrogen was carried out at 77 K. The specific surface areas of samples were evaluated by means of the Brunauer, Emmett and Teller (BET) theory ³⁰. Microporous and mesoporous volumes were determined by t-plot method of De Boer and Dubinin-Radushkevich equation respectively ^{31, 32, 33}.

X-Ray Diffraction (XRD) patterns were recorded at room temperature on a Siemens D5005 diffractometer using Cu Kα monochromatic radiation with 0.154050 nm wavelength. The samples were scanned in the 3 - 70° (2θ) range with a 0.01°/s step.

For the adsorption experiments in fixed-bed dynamic, the total gas flow at the entry and in exit of the reactor is fixed at 150 mL.min^-1. The concentration of each gas (CO₂, CH₄ and C₃H₈) is of 900 ppm and the gases flow is of 14 mL.min^-1 for each gas in order to obtain a concentration applied in entry of the reactor of 84 ppm. To obtain the fixed total flow (150 mL.min^-1), any difference is completed by a reconstituted air flow. Adsorption experiments were carried out at room temperature (295 K) with a bed height of 2.9 cm (550-650 mg of adsorbents).

Desorption experiments were carried out on the same bed of adsorbent, once this one was saturated. The principle is based on the sending of air reconstituted in entry of the reactor (flow: 80 mL.min^-1) with a simultaneous heating of the sample following a ramp of 24 K.min^-1 until 623 K. The desorbed compounds are analyzed on line using the chromatograph in gas phase and the gaseous composition in exit of the reactor.

3. Theoretical Aspects

The composition of the gas mixture outflow of the adsorbents bed is analyzed every four minutes. It is thus possible to plot the curve of drilling corresponding to each gas. These curves translate the efficiency of the bed to trap or adsorb a gas. The saturation of bed is reached when the concentrations at exit of the reactor (C) are equal to the values of the concentrations of the “blank” (C₀) that is the ratio C/C₀ equal to the unit. We obtain surface areas in minute, from integration by trapeze method:

C₀ and C_i are the concentrations of gas at the entry and outflow of the reactor at time T_i, respectively. In addition, the quantity of gas adsorbed (q) in the bed is given by relation:

where D, C_i, P, R, T and A_i correspond to the total flow of the gas mixture (mL min^-1), the concentration of the adsorbate (I) in entry of the reactor (ppm), the atmospheric pressure (bar), the perfect gas constant (J K mol^-1), the temperature (K) and the surface areas corresponding to the adsorbed quantity (min), respectively.

4. Results and Discussion

The discussions on the results of characterizations of studied adsorbents (BaX and NaX) were detailed in a recent article on the study of preferential adsorption sites of H₂O on adsorption sites of CO₂ at low temperature onto NaX and BaX zeolites ²⁸.

CO₂ adsorption from gas mixtures (CO₂/CH₄ and CO₂/C₃H₈) on NaX and BaX zeolites was carried out in fixed-bed dynamic to study the influence of molecules of neighboring size and molecular weight of carbon dioxide (CH₄ and C₃H₈, respectively) on the CO₂ adsorption capacities and the diffusion of the CO₂ molecule into the pores of the adsorbents. The adsorption capacities of methane and propane were also determined to evaluate the pore diffusion and affinity of gases with the studied adsorbents.

Table 1 shows the amounts of the adsorbed gases (CO₂ and CH₄) of gas mixture CO₂/CH₄ obtained on NaX and BaX zeolites. Those obtained during the adsorption of carbon dioxide only (CO_{2 ref}.), obtained in a recent study are also listed as reference ²⁸. The amounts of CO₂ adsorbed before breakthrough and at saturation on NaX zeolite (0.128 and 0.144 mmol.g^-1 respectively) are lower than the amounts determined in the case of the adsorption of CO₂ only (Table 1).

The NaX zeolite loses about 17% of the CO₂ adsorption capacities in the presence of CH₄. Methane seems to limit the accumulation of carbon dioxide in the pores of the NaX zeolite and / or limit the CO₂/cation interactions since the CO₂ adsorption takes place principally on cationic sites.

Unlike NaX, BaX zeolite increases the CO₂ adsorption capacities of 18% in presence of CH₄. Indeed, the amounts of CO₂ adsorbed obtained in adsorption of binary mixture before breakthrough and at saturation (0.078 and 0.110 mmol.g^-1 respectively) are higher than those obtained in CO₂ adsorption only (0.067 and 0.093 mmol.g^-1 respectively) as shown in Table 1. The presence of CH₄ seems to improve the accumulation of the CO₂ molecule in the pores of BaX zeolite.

The capacities of methane adsorption are practically similar for both adsorbents (Table 1). However, compared to the adsorption capacities of CO₂, CH₄ is not adsorbed on adsorbents studied. In fact, the quantities of CH₄ adsorbed are very small or negligible. These values are in the range of the experimental error (3.10^-3 mmol.g^-1).Thus, BaX and NaX zeolites show no affinity with methane in our experimental conditions. The recent research of Supatsorn Parinyakit et al. also indicated a low affinity of methane with 13X zeolite during the adsorption of the binary gas mixture CO₂/CH₄ ³⁴.

The Figure 1 shows the breakthrough curves in the CO₂/CH₄ binary mixture obtained on NaX zeolite. The breakthrough curve obtained in CO₂ adsorption only is also reported as reference. As expected, CH₄ has no affinity with NaX zeolite. The breakthrough is immediate on this zeolite.

The CO₂ breakthrough time obtained on NaX zeolite for the breakthrough curves of CO₂ adsorption in binary mixture is nearly similar compared to that obtained with CO₂ adsorption only (Figure 1). However, the slopes of breakthrough curves obtained when the bed is saturated are slightly different. This slight variation could indicate a slower diffusion of carbon dioxide molecule in the pores of NaX zeolite in the presence of methane or the weak interactions between CO₂ and cationic sites. The decrease in the slope of the CO₂ adsorption breakthrough curve in binary mixture confirms these observations. Indeed, the presence of methane decreases the capacities of CO₂ adsorption and modifies the diffusion of carbon dioxide in the porosity of the NaX zeolite.

The CO₂/CH₄ binary mixture breakthrough curves obtained on BaX are shown in Figure 2. The breakthrough curve obtained in CO₂ adsorption only is also reported as reference. As shown by the capacity at saturation, the breakthrough of methane is very fast as in the experiment with NaX zeolite.

The breakthrough time of CO₂ from binary mixture is higher than that obtained for CO₂ alone (Figure 2). As opposed to NaX zeolite, the slopes of the breakthrough curves are similar in both cases. The diffusion of CO₂ into the pores of BaX zeolite does not seems to be modified in the presence of CH₄. Thus, the presence of methane increases the adsorption capacities on BaX zeolite (Table 1) and has no effect on the diffusion of carbon dioxide into the pores of this adsorbent.

Table 2 lists the capacities of adsorption of CO₂ and C₃H₈ obtained on the NaX and BaX zeolites. Contrary to methane, propane is adsorbed by the adsorbents studied. There are thus adsorbate-adsorbent interactions between these materials and propane.

The adsorbed C₃H₈ amounts before breakthrough and at saturation obtained on NaX zeolite are 0.007 and 0.035 mmol.g^-1 respectively. These amounts of C₃H₈ adsorbed on NaX are lower compared to those obtained on BaX zeolite before breakthrough (0.035 mmol.g^-1) and at saturation (0.072 mmol.g^-1). It seems as on BaX zeolite, the propane/adsorbent interactions are more important.

As in presence of CH₄, the adsorbed CO₂ amounts before breakthrough and at saturation obtained on NaX (0.104 and 0.115 mmol.g^-1 respectively) are lower than those obtained when trapping CO₂ alone. The percentage loss of CO₂ adsorption capacity in presence of C₃H₈ is estimated to be around 30%. It thus appears that on NaX zeolite, the presence of a molecule with a molar mass close to CO₂ affects more strongly the capacities of CO₂ adsorption.

However on BaX zeolite, the adsorbed CO₂ amounts before breakthrough (0.079 mmol.g^-1 and at saturation (0.094 mmol.g^-1) are similar to those obtained with methane (Table 1). As in the case of methane, the presence of propane improves the accumulation of CO₂ in the porosity of BaX zeolite in our experimental conditions. In addition, the adsorbed CO₂ amounts are near to those obtained on the NaX zeolite at saturation (Table 2).

The breakthrough curves of CO₂ and C₃H₈ obtained on NaX zeolite are shown in Figure 3. For comparison, curve obtained when trapping CO₂ alone (CO_{2 ref}) is also shown. The breakthrough curve of C₃H₈ shows a steep slope, indicating weak interactions between C₃H₈ and NaX zeolite and / or rapid diffusion of the methane molecule into the porosity of the NaX faujasite.

Furthermore, no desorption of C₃H₈ due to CO₂ is observed. This means that there is no competitive adsorption between propane and carbon dioxide. However, NaX zeolite remains more efficient for the capture of carbon dioxide in a binary mixture CO₂/C₃H₈ in our experimental conditions (Table 2).

Nevertheless, a decrease in the breakthrough time for CO₂ in the presence of C₃H₈ is visible by comparing with the breakthrough curve of CO₂ adsorption alone (Figure 3). Moreover, the slopes of the CO₂ breakthrough curves obtained in binary mixture or alone are similar. This indicates that the presence of propane does not affect the diffusion of CO₂ in the porosity of the NaX zeolite. Thus, the presence of propane during CO₂ trapping decreases only the efficiency of the NaX zeolite.

The BaX zeolite, on the other hand, shows a strong adsorption of C₃H₈ with strong interactions between propane and the zeolite structure as shown in Figure 4. Indeed, the breakthrough time of C₃H₈ (50 min) is higher than that obtained on the NaX faujasite (8 min). The adsorption capacities calculated confirm these results (Table 2). These observations suggest that there are more favorable sites for C₃H₈ adsorption on BaX adsorbent.

However, as opposed to NaX zeolite, a slight desorption of propane is visible (C/C₀ > 1). There is thus a competitive adsorption between carbon dioxide and propane on the BaX zeolite (Figure 4). It seems that on BaX zeolite, CO₂ has preferential adsorption sites that C₃H₈ cannot occupy. Nevertheless, the presence of propane does not disturb the capacities of saturation adsorption of carbon dioxide in the BaX adsorbent. Also contrary to CH₄ molecule, in the presence of C₃H₈ molecule, the slope of CO₂ breakthrough curve for adsorption of CO₂/C₃H₈ mixture (Figure 4) is different from that obtained when trapping CO₂ alone (CO_{2 ref}). This reflects a more rapid diffusion of CO₂ in the porosity.

Thus, BaX zeolite can be used in selective adsorption or gas separation processes of C₃H₈ from gas effluents containing C₃H₈ with low gas concentrations. Indeed, this adsorbent presents favorable adsorption sites for propane. This result needs to be confirmed in the desorption study of CO₂ and C₃H₈.

Desorption of the gases, once the beds were saturated, was carried out to study the effect of CH₄ and C₃H₈ molecules on the preferential adsorption sites of carbon dioxide, mainly at low temperature (295 – 623 K) on NaX and BaX zeolites. The affinity of the gases with the adsorbents was also evaluated.

Figure 5 shows the evolution of the percentage of desorbed CO₂ and CH₄ obtained when the temperature increases obtained on NaX zeolite during the adsorption of the binary mixture CO₂/CH₄. For comparison, that obtained when trapping CO₂ alone (CO_{2 ref}) is also shown.

No CH₄ is observed in the desorption profiles obtained. Methane is therefore not adsorbed in the NaX zeolite, as shown by the adsorption capacities calculated (Table 1).

The CO₂ desorption curve in binary mixture shows the same distribution of adsorption sites as that obtained when the CO₂ was desorbed alone (Figure 5). On the other hand, a temperature shift of the desorption profile in the range 295 – 483 K is observed. This corresponds to a loss of the preferential CO₂ adsorption sites between 463 - 483 K. This result could be explained by the fact that methane limits the accessibility of CO₂ to these sites. On the other hand, at high temperatures (573 – 623 K), the quantities desorbed remain similar (Figure 5). According to the results, in NaX zeolite, the presence of CH₄ molecule does not modify the proportions of desorbed CO₂. However, the temperature distribution of the physisorbed carbon dioxide species is decreased.

On the BaX zeolite, the desorption curves obtained are showed in Figure 6. As with NaX, methane is not detected during the desorption. The quantities adsorbed are indeed very small (Table 1).

The desorption curves of CO₂ in mixture or alone (CO_{2 ref}) are similar (Figure 6). No modification of desorbed species proportions and temperature shift in the distribution of CO2 preferential adsorption sites is observed in the range of temperatures explored. This confirms the capacity of the breakthrough curves. Methane does not disturb the distribution of CO₂ adsorption sites and does not interfere with the access of carbon dioxide to its preferential adsorption sites on the BaX zeolite.

Desorption curves of carbon dioxide and propane obtained on NaX and BaX are shown in Figure 7 and Figure 8. The CO₂ desorption profile obtained when trapping CO₂ alone (CO_{2 ref}) is also shown.

The curve obtained on the NaX zeolite shows a desorption of C₃H₈ in the temperature range between 295 – 333 K with a maximum of desorption around 323 K (Figure 7). There are thus, on the zeolite NaX, sites favorable to the adsorption of propane.

The desorption curve for CO₂ in the CO₂/C₃H₈ binary mixture (Figure 7) shows a distribution of desorbed CO₂ percentages similar to that obtained for CO₂ alone. As with the CH₄ profile (Figure 5), an offset (283 K) of temperature of the desorption maximum is observable. This temperature corresponds to the end of desorption of C₃H₈. It appears thus that CO₂ occupies different adsorption sites than propane. A decrease in the proportions of desorbed species (CO₂) between 343 - 483 K is also observed. Therefore, the loss of CO₂ capacity obtained in the presence of C₃H₈ could be attributed to the decrease of trapped species in this temperature range. Propane has thus made some sites inaccessible to CO₂.

For BaX zeolite (Figure 8), the curves of CO₂ and C₃H₈ show a desorption in the same temperature range (295 – 373 K). This result confirms that methane interacts more strongly with the BaX zeolite than with NaX. Indeed, the temperature at the desorption maximum is 356 K with a total desorption around 373 K. As in the case of methane, the distribution of preferential adsorption sites of CO₂ in the presence of methane does not appear to be modified on BaX (Figure 8). However, between 423 – 532 K, a weak desorption of CO₂ is still observable. These species are probably adsorbed on strong sites. The presence of C₃H₈ does not therefore disturb the access of carbon dioxide to these preferential adsorption sites.

5. Conclusion

The dynamic adsorption and desorption of carbon dioxide (CO₂) in gas mixtures with methane (CH₄) and propane (C₃H₈) at low gas concentration (84 ppm) was carried out on BaX and NaX zeolites in order to study the influence of molecules with sizes and molecular weights similar to those of carbon dioxide (CH₄ and C₃H₈ respectively) on: the capacities of carbon dioxide adsorption, the diffusion of CO₂ into the porosity of the adsorbents and/or the CO₂-adsorbent interactions and on the adsorption sites preferential of CO₂ in the temperature range of physisorbed carbon dioxide (295 - 623 K). The affinity and adsorption capacities of CH₄ and C₃H₈ on the adsorbents was also evaluated.

The capacities of CH₄ at saturation obtained on NaX and BaX zeolites (0.004 and 0.003 mmol.g^-1 respectively) are insignificant. It seems that CH₄ molecule has no affinity with NaX and BaX zeolites in our experimental conditions. The breakthrough and desorption curves of CH₄ obtained confirm the conclusions. On the other hand, whatever the adsorbent studied, the results obtained show favorable adsorption sites to the trapping of C₃H₈ molecule. The adsorption capacities of propane at saturation on NaX and BaX zeolites are 0.035 and 0.075 mmol.g^-1 respectively. It appears that BaX zeolite is the adsorbent which presents more affinity for C₃H₈ molecule and the strongest propane/zeolite interactions. Indeed, the exploitation of the desorption curves of C₃H₈ reveals on BaX, the temperatures of the maximum desorption peak (356 K) and total desorption of propane (373 K) higher than those obtained on NaX zeolite (320 and 338 K respectively).

The results of CO₂ adsorption in the presence of methane and propane on NaX zeolite indicate a decrease in the CO₂ / cation interactions, marked by a loss of capacities of about 17% and 30% when carbon dioxide trapping is carried out in binary mixture CO₂/CH₄ and CO₂/C₃H₈ respectively. Moreover, in the presence of methane, the CO₂ desorption study shows a loss of preferential adsorption sites of CO₂ between 443 - 483 K.

However on BaX zeolite, the presence of CH₄ and C₃H₈ improves the accumulation of CO₂ in the porosity. Indeed, the capacities of CO₂ adsorption are improved of about 18% whatever the binary mixture of gas studied. Though, the diffusion of CO₂ in the porosity of BaX zeolite is modified only in the presence of C₃H₈.

Moreover, as opposed to NaX zeolite, the CO₂ desorption study indicates a conservation of preferential adsorption sites of CO₂ on BaX zeolite in the temperature range between 295 - 623 K, whatever the gas effluent studied.

Therefore, in the presence of methane and propane, BaX zeolite keeps all its properties in the selective adsorption of carbon dioxide in this study. Thus, BaX zeolite can be used in the selective adsorption (gas separation) or valorization processes of C₃H₈ from effluents with low gas concentrations. Indeed, it presents favorable adsorption sites for propane with low total desorption temperatures that are advantageous for adsorbent regeneration treatments.

Acknowledgments

This study was financially supported by the Research Program of the Université des Sciences et Techniques de Masuku and the Université de Poitiers.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References