Chinese Chemical Letters  2016, Vol. 27 Issue (6): 933-937   PDF    
Separation of CO2–N2 using zeolite NaKA with high selectivity
Bo Yang, Yan Liu, Ming Li     
Department of Chemistry, Tongji University, Shanghai 200092, China
Abstract: The adsorption method based on solid adsorbents is one of feasible ways to capture and store CO2. Using the ion exchange method, different zeolites NaKA varying in K+ content were produced. The adsorption isotherms and kinetic uptakes were measured. The experimental results show that the optimal NaKA could adsorb significant quantities of CO2 and little N2. On the zeolite NaKA with 14.7 at.% K+, the adsorption capacity for pure CO2 is over 3.10 mmol g-1 and the CO2-N2 selectivity is about 149 at ambient pressure and temperature. The kinetic CO2-N2 selectivity could also achieved 200 within 3 min according to the uptake data. To demonstrate the separation effectiveness, breakthrough curves of pure components and binary mixtures were investigated experimentally and theoretically in a fixed bed. It is found that the breakthrough points of CO2 and N2 are almost at the same time under the atmospheric pressure at 348 K with the raw gas composition CO2/N2 (20:80, v/v). If the pressure has been increased higher than 0.1 MPa, CO2 would break through the bed much slower than N2. Therefore, the pressure may become the limiting factor for the separation performance of zeolites NaKA.
Key words: NaKA zeolite     Selectivity     Adsorption     CO2 capture     Separation    
1. Introduction

CO2 capture and storage (CCS) from flue gas,as a vital topic in regard to environmental protection,has already been widely concerned [1]. Since the main object for the CCS is to separate CO2 from N2,the adsorption method based on solid adsorbents has been considered to reduce the high cost of the gas separation process,comparing with other technologies such as cryogenic rectification,selective membranes and solvent absorption techniques [2]. Due to the pressure of flue gas close to atmosphere and the low concentration of CO2,it is of great importance to improve the efficiency of CCS by adsorption separation. Thus many explores have been made recently in the development of high-performance adsorbent materials.

Most promising adsorbents for CCS could be classified as three categories: organics,organic-inorganics and inorganics. The porous metal-organic frameworks (MOFs) display excellent performance in extracting CO2 from a gas mixture after chemical modification [3-5]. As organic-inorganic hybrid adsorbents,solid amine [6] selectively adsorbs CO2 through chemisorption and ZIF- 78 (zeolitic imidazole frameworks) [7] has an aperture of 3.8A˚ which is also meaningful to improve the CO2-N2 selectivity. However,it should be noted that these two kinds of adsorbents are not capable of CCS for large-scale industrial applications due to the instability and high cost. Consequently,inorganic adsorbents including metal oxides,hydrotalcites,carbon and zeolites [8, 9] may be more promising for industrial usage. At present,zeolites have been widely applied because of their unique ability of molecular sieving. There are several zeolites of commercial interest,such as type A,X,Y and ZSM [10]. Even though zeolite 13- has the highest adsorption capacity (0.7 mmol g-1,101 kPa,393 K) among five zeolites (4A,5A,13-,APG-II and WE-G 592) [11],a recent research [12] has illustrated that a zeolite NaKA with 17 at.% k+ displayed a high ideal CO2-N2 selectivity as 172 with CO2 capacity as 3.43 mmol g-1 at 298 K and 85 kPa. Another work [13] also showed that a very high IAST CO2-N2 selectivity (>1100,298 K,101 kPa) could be obtained from a monolith of zeolite NaKA with 9.9 at.% k+. In addition,Cheung et al. [14] has reported that the zeolite ZK-4 had the same structure as zeolite A and the zeolite NaK-ZK-4 (26 at.% extraframework cations of ZK-4 exchanged by k+) adsorbed a large amount of CO2 (4.35 mmol g-1) but negligible N2 (<0.03 mmol g-1) with 273 K and 101 kPa. It could be followed that the zeolite NaKA could potentially offer a high selectivity performance superior to other inorganic adsorbents.

In this work,five kinds of zeolite NaKA varying in k+ content were synthesized using the ion exchange method. The adsorption isotherms and kinetic uptakes (CO2 and N2) on these zeolites were measured. Considering that the separation performance of the NaKA zeolites in an adsorption bed has not been demonstrated yet,attempts have been made to further investigate the separation effectiveness in this study. Consequently,breakthrough curves of pure components and binary mixtures were measured in a fixed bed. In addition,theoretical calculations were performed to simulate and analyze the breakthrough curves,in order to identify the important factors in the adsorption separation process.

2. Experimental 2.1. Materials

The powder of NaA zeolite was provided by Shanghai Zeolite Molecular Sieve Co.,Ltd.,China. The analytically pure KCl (China National Medicines Co.,Ltd.) was used for the ion exchange process. Pure CO2,N2 and He with purity above 99.99% were used. The gas mixture was made by pure components flowing through the mass flow controllers (Seven Star D07,China) before into a cylinder mixer. Analyses of gases were performed using the TCD detector of a gas chromatography (Techcomp 7890II,China).

To prepare the NaKA zeolites with different degree of k+ exchange,the concentration of KCl was varied from 0.025 mol L-1 to 0.500 mol L-1. Detailed preparing process could be referenced from the article [12]. Then,the ion concentration of the solution after finishing the exchange process was measured by inductively coupled plasma optical emission spectrometry (Optima 2100 DV,PerkinElmer). The results were substituted to the balance:

$NaA+{{K}.{+}}\Leftrightarrow NaKA+N{{a}.{+}}$ (1)

Therefore,the k+ ion content of zeolite NaKA could be determined through the material balance. Then five kinds of zeolite NaKA powder with different k+ ion content were obtained.

2.2. Adsorption isotherms and kinetics

Adsorption equilibrium isotherms of pure CO2 and N2 on the NaKA zeolites were measured at 293 K by the volumetric method (Fig. 1). Details of the experimental apparatus and experimental procedures have been illustrated in the previous work [15]. The adsorption data for CO2 and N2 were fitted to the Langmuir equation. The parameters of b and qsi are given in Table 1.

$\frac{q_{i}.{*}}{{{q}_{si}}}=\frac{{{b}_{i}}{{p}_{i}}}{1+{{b}_{i}}{{p}_{i}}}$ (2)
Download:
Figure 1. Adsorption isotherms of pure gas fitted by the Langmuir model on the zeolites NaKA at 293 K. Scatters: experimental data; lines: the Langmuir model ((a) CO2; (b) N2).

Table 1
Langmuir constants of CO2 and N2 on the zeolites NaKA.

The uptake rates of each gas were also measured at 0.1 MPa and 293 K on the five kinds of NaKA zeolites. The experimental results were shown in Fig. 2. When an optimal adsorbent was chosen among the five zeolites,the effectiveness to separate the CO2/N2 mixture has been investigated under 348 K,due to the temperature of flue gas generally ranging from 343 K to 423 K. The adsorption isotherms of the chosen adsorbent were also measured at 313 K and 348 K,as depicted in Fig. 3. The Langmuir parameters are also listed in Table 1. The pure and binary gas breakthrough curves were measured in a single-bed at 348 K. The adsorption bed used in this experiment was a stainless steel tube with the length of 180 mmand the internal diameter of 18 mm. The adsorbents were the cylindrical zeolite NaKA pellets of 1.4 mmin diameter. For pure gas breakthrough experiments,the feed flow rate through the bed was 40 mL min-1 with the pressure 0.45 MPa. The helium was adopted to pre-pressurize the bed. In addition,for binary mixture breakthrough experiments,the feed flow rate was 60 mL min-1 and the pressure was 0.3 MPa. The composition of feed gas was varied as 10/90,20/80 and 50/50 vol% CO2/N2. Other experimental details could be found in Table 2. The experimental breakthrough curves are illustrated in Figs. 4 and 5.

Download:
Figure 2. Kinetic uptakes of pure gas on the zeolites NaKA at 0.1 MPa and 293 K ((a) CO2; (b) N2).

Download:
Figure 3. Adsorption isotherms of pure CO2 and N2 fitted by the Langmuir model on the zeolites NaKA (14.7 at.% k+) at 293, 313 and 348 K. Scatters: experimental data; lines: the Langmuir model.

Table 2
Experimental information and parameters used in the simulation.

Download:
Figure 4. Breakthrough curves of pure CO2 and N2 in a single bed. Experimental conditions: P = 0.45 MPa; T = 348 K; flowrate of raw gas = 40 mL min-1; adsorbents: the zeolite NaKA with 14.7 at.% k+ ion content. Scatters: experimental data; lines: simulation.

Download:
Figure 5. Breakthrough curves of binary mixtures in a single bed. Experimental conditions: P = 0.3 MPa; T = 348 K; flowrate of raw gas = 60 mL min-1; composition of raw gas: 10/90, 20/80, 50/50 vol% CO2/N2; adsorbents: the zeolite NaKA with 14.7 at.% k+ ion content. Scatters: experimental data.

2.3. Mathematical model

Assuming the ideal gas law for each gas,isothermal behavior and the axisymmetric plug flow,the following models could be obtained to describe the breakthrough processes.

The gas-phase mass balance for each component can be expressed as [16]

$-{{D}_{L}}\frac{{{\partial }.{2}}{{c}_{i}}}{\partial {{z}.{2}}}+{{c}_{i}}\frac{\partial v}{\partial z}+v\frac{\partial {{c}_{i}}}{\partial z}+\frac{\partial {{c}_{i}}}{\partial t}+\frac{1-{{\varepsilon }_{b}}}{{{\varepsilon }_{b}}}{{\rho }_{b}}\frac{\partial {{{\bar{q}}}_{i}}}{\partial t}=0$ (3)

where DL is the axial diffusion coefficient,ci and qi are the concentration of component i in the gas phase and solid phase,respectively,v is the gas velocity inside the bed,z is the distance along the bed and t is the time. eb represents the inter-particle voidage while rb symbolizes the packed bed bulk density. The linear driving force (LDF) model [16] has been adopted to describe the mass transfer process

$\begin{align} & -{{D}_{L}}\frac{{{\partial }.{2}}{{c}_{i}}}{\partial {{z}.{2}}}+{{c}_{i}}\frac{\partial v}{\partial z}+v\frac{\partial {{c}_{i}}}{\partial z}+\frac{\partial {{c}_{i}}}{\partial t}+\frac{1-{{\varepsilon }_{b}}}{{{\varepsilon }_{b}}}{{\rho }_{b}}\frac{\partial {{{\bar{q}}}_{i}}}{\partial t}=0 \\ & \\ & \frac{\partial {{{\bar{q}}}_{i}}}{\partial t}={{k}_{i}}(q_{i}.{*}-{{{\bar{q}}}_{i}}) \\ \end{align}$ (4)

where ki could be attained by fitting the experimental data of pure component breakthrough curves,as listed in Table 2. The equilibrium amount,qi*,of binary mixtures is approximated using the extended Langmuir equation

$\frac{q_{i}.{*}}{{{q}_{si}}}=\frac{{{b}_{i}}{{p}_{i}}}{1+\sum{{{b}_{i}}{{p}_{i}}}}$ (5)

Based on Eq. (3), the overall mass balance with constant pressure takes the form

$\frac{\partial v}{\partial z}+\frac{RT}{P}\frac{1-{{\varepsilon }_{b}}}{{{\varepsilon }_{b}}}{{\rho }_{b}}\sum{\frac{\partial {{{\bar{q}}}_{i}}}{\partial t}}=0$ (6)

The above partial differential equations have been solved by the orthogonal collocation method after being transformed in the dimensionless form. Sixteen internal collocation points are adopted with the relative tolerance less than 10-6.

3. Results and discussion 3.1. Adsorption isotherms and kinetic uptakes

The experimental and calculated adsorption isotherms of CO2 and N2 on five kinds of zeolite NaKA at 293 K are depicted in Fig. 1. The experimental data could be predicted favorably by the Langmuir model. It is obvious that both the adsorption capacities of CO2 and N2 decreases as the k+ ion content increases. It could be ascribed to that the pore size of zeolite NaA (3.8A˚ ) narrowed because of the k+ replacing Na+. It would become harder for the gas molecules of CO2 and N2 to get into the micropores. The investigation also indicates that the adsorption capacity of CO2 under 0.1 MPa has achieved to 3.55 mmol g-1 while that of N2 is only 0.22 mmol g-1 on the zeolite NaKA with 10.2 at.% k+ content. When the k+ ion content increases to 14.7 at.%,the zeolite NaKA adsorbs almost no N2 but the amount of adsorbed CO2 just reduces slightly to 3.10 mmol g-1. If the k+ ion content continued increasing,it should be noted that the equilibrated amount of CO2 would decrease greatly. In addition,the zeolite NaKA scarcely adsorbs N2 when the k+ ion content is larger than 22.5 at.%,which causes the lack of the Langmuir parameters for N2 in Table 1.

The kinetic uptakes of both CO2 and N2 on the zeolites NaKA under 0.1 MPa and 293 K are presented in Fig. 2. It appears that if the k+ exchange for zeolite NaA is more,the uptake rates of gas molecules are slower. It can also be seen that the uptake rates of CO2 remain in a high level during the first ten minutes,as illustrated in Fig. 2(a). For the zeolitesNaKAwith 10.2 at.%,12.8 at.%,and 14.7 at.% k+,the adsorption amounts could approach 2.07,1.85,1.80mmol g-1,respectively,within twenty minutes. But for the other two zeolitesNaKAwith22.5 at.%and 29.3 at.%k+,theamounts of adsorption are only 1.36 and 0.54mmol g-1,respectively. In contrast,it ismuch faster forN2 to reach the adsorption equilibrium,as shown in Fig. 2(b). For the zeolite NaKA with 10.2 at.% k+,it takes over 80% of the maximum uptake capacity for N2 within five minutes,which is about 0.13mmol g-1. In addition,not only are the uptake rates for the other four zeolitesNaKAmore rapid,but also the adsorbed amounts are quite small. Thus it would be possible to control the time for adsorption processes in the range of 5 min to 20 min to obtain satisfactory separation effectiveness.

On the basis of the above analyses,it is obviously promising to recover CO2 from N2 using the zeolite NaKA for its steric mechanism. It could also be concluded that the selectivity would vary according to both the degree of the k+ ion exchange and the adsorption time.

3.2. Selectivity

To discuss the CO2-N2 selectivity on different zeolites NaKA in detail,the selectivity,Se,based on their isotherms is defined as [17]

${{S}_{e}}=\frac{{{x}_{c{{o}_{2}}}}/{{y}_{c{{o}_{2}}}}}{{{x}_{{{N}_{2}}}}/{{y}_{{{N}_{2}}}}}$ (7)

where xi (i = CO2,N2) represents the adsorbed fraction in the adsorbate phase,which were predicted by the IAST model [18],and yi is the ratio of mole fraction in the bulk gas phase. It has been presented in Fig. 6(a) of the results of Se for a gaseous mixture with a composition of 20/80 vol% CO2/N2. Apparently,the IAST selectivity increaseswith the gas pressure rising. Especially for the zeoliteNaKA (14.7 at.% k+),the selectivity has a great increasing tendency. It also indicates that the Se of zeolites NaKA with 10.2 and 12.8 at.% k+ are much smaller than those of zeoliteNaKAwith14.7 at.% k+,which are about 2,5 and 149 at 0.1 MPa,respectively. The results of IAST selectivity on zeolite NaKA (14.7 at.% k+) reach a high level (larger than 100) in consistent with the literatures [12-14],which is substantial to the CCS from flue gas. Even though there are deviations for the data of selectivity,it could be ascribed to that the low adsorption amount of N2 leads to more measurement errors. It should also be noticed that the IAST selectivity of zeolites NaKA (22.5 at.% and 29.3 at.% k+) is unavailable in the figure due to the absence of the Langmuir constants of N2.

Download:
Figure 6. The CO2–N2 selectivity on the zeolites NaKA ((a) IAST selectivity; (b) Kinetic selectivity).

Since the adsorption time is limited as discussed in Section 3.1,a desirable adsorbent should possess the ability to achieve high CO2 capacities within a short time and to maintain a high CO2-N2 selectivity simultaneously. Considering that,it has been investigated of the relationships between the kinetic selectivity,Sk,and the adsorption time on different zeolites at 0.1 MPa and 293 K,which have been illustrated in Fig. 6(b). The Sk is calculated by using the instantaneously adsorbed amount of CO2 divided by that of N2,both of which have already been presented in Fig. 2. It could be found that the Sk on zeolites NaKA with k+ ion content below 12.8 at.% change little along with the time increasing and remain under 50. The Sk for 29.3 at.% k+ has raised to a high level as over 800 within five minutes. If the adsorption capacity of CO2 on this zeolite NaKA (29.3 at.% k+) could be greater,it would be a choice of candidate adsorbents for separating CO2-N2 mixtures. Furthermore,the zeolite NaKA with 22.5 at.% k+ stands in a similar situation to the zeolite NaKA (29.3 at.% k+). In particular,the kinetic CO2-N2 selectivity on the zeolite NaKA (14.7 at.% k+) could achieve over 200 within three minutes. Therefore,it could regard the zeolite NaKA (14.7 at.% k+) as the proper adsorbent since it has a balance of the apposite adsorption capacity,the rapid uptake rate and the high selectivity.

3.3. Separation effectiveness

The experimental and simulated breakthrough curves of pure gas on zeolite NaKA with 14.7 at.% k+ ion content are illustrated in Fig. 4. It seems that N2 could flow through the adsorption bed more quickly than CO2 at 0.45 MPa and 348 K. The constants ki in the linear driving force model for CO2 and N2,have been determined as 9.6 - 10-4 s-1 and 3.6 - 10-4 s-1,respectively,listed in Table 2. It can be seen that CO2 could diffuse more easily into the adsorbent micropores,since the CO2 molecule has the smaller kinetic diameter. It also proves that controlling the degree of k+ exchange for NaA to adjusting the pore size could be an effective method to obtain an optimum adsorbent for the CO2-N2 separation.

The experimental and predicted breakthrough curves of the CO2/N2 mixture at 0.3 MPa and 348 K are presented in Fig. 5. It is foreseeable that N2 has flowed through the adsorption bed quite quickly,even though its concentration in raw gas has reduced to 50 vol%. Although the breakthrough time for CO2 varied little,its relative concentration (C/C0,C and C0 represent the concentration at the outlet and inlet of the adsorption bed,respectively) at the outlet has decreased as the primary CO2 concentration raised to 50 vol% from 10 vol%. The reason could be attributed to that the partial pressure of CO2 is improved in consideration of that the total pressure of rawgas remains unchanged. It seems that the separation of CO2/N2 is not good enough based on the zeolite NaKA,despite the high selectivity. It could be explained that the partial pressure of CO2 is usually low in the flue gas,which cannot provide high adsorption capacity. Naturally,if keeping the composition of raw gas constant but raising the gas pressure,the partial pressure of CO2 will also be raised and then the separation effectiveness for the CO2-N2 mixture may be refined. To verify this viewpoint,the predicted binary breakthroughcurves inFig. 7 areobtained.When the pressure ofraw gas rises to 1.5MPa,the breakthrough time of N2 has hardly changed. However,the time for CO2 to break through the bed has been well prolonged. Thus it is evident that high pressure is benefit to the improvement of CO2-N2 separation performance. The predicted breakthrough curves at 0.3 MPa have been compared with the experimental data as shown in Fig. 7,which demonstrates that the calculated results are capable of representing breakthrough curves favorably. However,it should be mentioned that rising raw gas pressure will cause additional cost for operation. Since the pressure of flue gas usually approaches to atmosphere,it is worth evaluating whether to take high pressure as the approach to realize efficient CCS when the zeolite NaKA has been adopted.

Download:
Figure 7. Predicted breakthrough curves of 20/80 vol% CO2/N2 mixture. Simulation conditions: P = 0.1, 0.3, 0.5, 1.0, 1.5 MPa; T = 348 K; flowrate of raw gas = 60 mL min-1; composition of raw gas: 20/80 vol% CO2/N2; adsorbents: the zeolite NaKA with 14.7 at.% k+ ion content. Scatters: experimental data; lines: simulation.

4. Conclusion

Five zeolites NaKA varying in k+ content were synthesized using the ion exchangemethod for the separation ofCO2-N2 mixtures. The adsorption isotherms,kinetic uptakes and breakthrough curves of CO2 and N2 were measured and analyzed experimentally and theoretically. It is attainable to obtain an optimized zeolite NaKA with high CO2-N2 selectivitywhich adsorbs almost little N2 with the significant amount of CO2. In this study,the zeolite NaKA with 14.7 at.% k+ content has been providedwith acceptable capabilities,including the adsorption capacity as 3.10mmol g-1,the IAST selectivity as 149 and the kinetic selectivity as over 200 within threeminutes. The adsorption breakthrough processes demonstrate the validity of NaKA for CO2/N2 separation with controlling the adsorption time and pressure. It should be mentioned that the separation effectiveness would be weakened by the low partial pressure of CO2 in the flue gas. Although raising the pressure could make the adsorption separation process executable,extra energy cost would be required.

References
[1] R.S. Haszeldine. Carbon capture and storage: how green can black be?. Science 325 (2009) 1647–1652.
[2] J.C.M. Pires, F.G. Martins, M.C.M. Alvim-Ferraz, et al. , Recent developments on carbon capture and storage: an overview. Chem. Eng. Res. Des. 89 (2011) 1446–1460.
[3] N. Casas, J. Schell, R. Blom, et al. , MOF and UiO-67/MCM-41 adsorbents for precombustion CO2 capture by PSA: breakthrough experiments and process design. Sep. Purif. Technol. 112 (2013) 34–48.
[4] R. Sabouni, H. Kazemian, S. Rohani. Mathematical modeling and experimental breakthrough curves of carbon dioxide adsorption on metal organic framework CPM-5. Environ. Sci. Technol. 47 (2013) 9372–9380.
[5] Q.Y. Yang, L.L. Ma, C.L. Zhong, et al. , Enhancement of CO2/N2 mixture separation using the thermodynamic stepped behavior of adsorption in metal-organic frameworks. J. Phys. Chem. C 115 (2011) 2790–2797.
[6] M.L. Chua, L. Shao, B.T. Low, et al. , Polyetheramine-polyhedral oligomeric silsesquioxane organic-inorganic hybrid membranes for CO2/H2 and CO2/N2 separation. J. Membr. Sci. 385-386 (2011) 40–48.
[7] R. Banerjee, H. Furukawa, D. Britt, et al. , Control of pore size and functionality in isoreticular zeolitic imidazolate frameworks and their carbon dioxide selective capture properties. J. Am. Chem. Soc. 131 (2009) 3875–3877.
[8] S. Choi, J.H. Drese, C.W. Jones. Adsorbent materials for carbon dioxide capture from large anthropogenic point sources. ChemSusChem 2 (2009) 796–854.
[9] N. Hedin, L.J. Chen, A. Laaksonen. Sorbents for CO2 capture from flue gas—aspects from materials and theoretical chemistry. Nanoscale 2 (2010) 1819–1841.
[10] S. Kulprathipanja. Zeolites in Industrial Separation and Catalysis. Wiley-VCH, New York, NY (2010) .
[11] R.V. Siriwardane, M.S. Shen, E.P. Fisher, et al. , Adsorption of CO2 on zeolites at moderate temperatures. Energy Fuels 19 (2005) 1153–1159.
[12] Q.L. Liu, A. Mace, Z. Bacsik, et al. , NaKA sorbents with high CO2-over-N2 selectivity and high capacity to adsorb CO2. Chem. Commun. 46 (2010) 4502–4504.
[13] F. Akhtar, Q.L. Liu, N. Hedin, et al. , Strong and binder free structured zeolite sorbents with very high CO2-over-N2 selectivities and high capacities to adsorb CO2 rapidly. Energy Environ. Sci. 5 (2012) 7664–7673.
[14] O. Cheung, Z. Bacsik, P. Krokidas, et al. , K+ exchanged zeolite ZK-4 as a highly selective sorbent for CO2. Langmuir 30 (2014) 9682–9690.
[15] M. Li, J. Liu, T.L. Wang. Adsorption equilibria of carbon dioxide and ethane on graphitized carbon black. J. Chem. Eng. Data 55 (2010) 4301–4305.
[16] D.M. Ruthven, S. Farooq, K.S. Knaevel, Pressure Swing Adsorption, Wiley-VCH, New York, N Y. , 1994. Wiley-VCH, New York, NY (1994) .
[17] R.T. Yang. Gas Separation by Adsorption Processes. Imperial College Press, London (1997) .
[18] A.L. Myers, J.M. Prausnitz. Thermodynamics of mixed-gas adsorption. AIChE J. 11 (1965) 121–127.