Desalination 193 (2006) 313–320
CO2 facilitated transport through an acrylamide and maleic anhydride copolymer membrane Li-guang Wua*, Jiang-nan Shenb, Huan-lin Chenb, Cong-jie Gaoa a
The Center for Development of Water Treatment Technology, State Ocean Administration, Hangzhou 310012, China Fax: +86 (571) 8886 8427; email:
[email protected] b College of Materials and Chemical Engineering, Zhejiang University, Hangzhou 310027, China Received 11 January 2005; accepted 7 April 2005
Abstract A copolymeric membrane material containing an amine group for CO2 facilitated transport was synthesized by radical polymerization of acrylamide and maleic anhydride. The relative molecular weight and chemical structure of the copolymer were analyzed by viscometric measurement, elemental analysis, and FTIR. The copolymeric membranes were prepared. The sorption behavior as well as the permeabilities of the membranes for pure CO2 and CH4 were investigated. The results show that the copolymeric membrane possesses a higher permeability of CO2 and a lower permeability of CH4. The membrane displays a CO2 permeability of 5×10!12 cm3(STP)/cm2.s.pa, and a CH4 permeability of 2×10!13 cm3(STP)/cm2.s.pa. CO2 sorption behavior of the copolymeric membrane, which can be classified as a dual-mode sorption model, and CH4 sorption behavior of the copolymeric membrane accord with the Fickian diffusion model. Keywords: Facilitated transport membrane; Sorption behavior; Permeability; CO2; CH4
1. Introduction CO2 removal from gas streams is required in many industrial applications among which natural gas processing is probably the most important one. Polymeric membranes are used to remove *Corresponding author.
CO2 from natural gas, but few commercially available polymeric membranes have both high permeability and high selectivity [1]. Facilitated transport membranes exhibit fairly high selectivity as well as high permeability because of carrier-mediated transportation. Facilitated transport membranes involve carrier-mediated transportation in addition to
Presented at the International Congress on Membranes and Membrane Processes (ICOM), Seoul, Korea, 21–26 August 2005. 0011-9164/06/$– See front matter © 2006 Published by Elsevier B.V.
doi:10.1016/j.desal.2005.04.146
314
L.-g. Wu et al. / Desalination 193 (2006) 313–320
permeant physical dissolution and diffusion. The existence of a carrier that can react reversibly with the permeant brings about high selectivity and high permeability. There are two types of facilitated transport membranes: one is a mobile carrier membrane (liquid membrane) where the carrier can diffuse in the membrane, and the other is a fixed carrier membrane where the carrier is immobilized in the membrane matrix and cannot move. Although there is a remarkably high selectivity for liquid membranes, the disadvantage of this type of membrane is that it has poor stability. The fixed carrier membrane, in which the carrier is chemically bonded, is more favorable in comparison with the liquid membrane because there is little chance of carrier leakage. The fixed carrier membrane, which removes acid gas from gas streams and separates organic–organic mixtures, has been investigated extensively in recent years [2–7]. In this paper, the membranes containing amine groups, which act as fixed carriers for CO2 facilitated transport, were prepared, and their chemical structure analyzed. In addition,the CO2, CH4 sorption behavior of the membranes and their CO2, CH4 permeability were investigated. 2. Experimental 2.1. Materials Acrylamide (AM. m.p. 83-84) and maleic anhydride (MA. m.p. 56) were recrystallized from reagent-grade benzene and dried in a vacuum at room temperature. Azobisisobutyronitrile (AIBN) was recrystallized from reagent-grade methanol prior to use. Deionized water was used. 2.2. Synthesis of AM–MA copolymer AM (7.108 g), MA (9.806 g) and benzene (250 ml) were introduced into a 300-ml threenecked flask equipped with a stirrer, a condenser
and a thermometer. The contents were stirred at 80 for complete dissolution. AIBN (0.20 g in four steps) was added to the dissolution as an initiator, and was stirred for 3 h at the same temperature in a nitrogen atmosphere. The polymer precipitated from the solution as polymerization continued, and was then completed. The precipitated product was washed with acetone several times and a hygroscopic white solid was obtained. The yield of copolymers was more than 60 wt%. The products were placed into a mixture of methanol/ water (1/1), which was a solvent for these two monomers and the AM homopolymer. The insoluble copolymer was removed by filtration and dried under vacuum to a constant weight at room temperature. A large volume of methanol (more than 1/10) was added to the filtrate to check that the AM homopolymer was absent. 2.3. Membrane preparation A known concentration of aqueous AM–MA copolymer solution was cast onto a glass plate and the PSF ultrafiltration membrane to form a homogeneous membrane and a composite membrane. The membranes were dried at room temperature for a day. The thickness of the membrane strips was measured on a micrometer with an accuracy of 0.1 µm. The thickness of the homogeneous membranes was 10 µm, and the top layer thickness of the composite membranes was 2 µm. 2.4. Viscometric measurements Viscometric measurements were carried out with an Ubbelohde viscometer. Viscosity data were calculated according to the Mark-Huggins equations: ηsp/C = [η] + kN [η]2 C where ηsp is the specific viscosity, C is the concentration of copolymer; [η] is the intrinsic
L.-g. Wu et al. / Desalination 193 (2006) 313–320
viscosity; and kN is the Huggins constant. The value of ηsp/C plotted vs. C, [η] and kN are, respectively, determined from the intercept and slope of the straight line. The viscosity–average molecular weight of the AM–MA copolymer was calculated from the intrinsic viscosity ([η]) according to the equation [8] [η] = 1.05×10!5 M0.66 2.5. Characterization techniques The composition of the AM–MA copolymer was determined by the method of elemental analysis. An EA1110 CHNS-O (Carlo Erba) was used to determine the carbon, hydrogen, and nitrogen content of the copolymer. The IR spectra of the AM–MA copolymer were recorded with a Shimadzu IR-47 spectrophotometer using KBr discs in the range of 500– 4000 cm!1. 2.6. Sorption experiments The schematic diagram of the experimental apparatus is shown in Fig. 1. A homogeneous membrane strip of an accurate weight was hung
315
on a quartz spring and loaded in the sorption chamber where the temperature of the chamber was kept constant. A vacuum was applied to the sorption chamber after pure N2 was repeatedly introduced and removed. When an extension of the quartz spring which was loaded with a strip was monitored to be time invariant, CH4 (or CO2) was introduced into the sorption chamber. A linear variation of the quartz spring due to the absorption of CH4 (or CO2) molecules was detected using a cathetometer to a precision of 10 µm. 2.7. Permeation experiments Steady-state gas permeability was determined for CO2 and CH4 using a permeration apparatus, which is shown in Fig. 2 as a schematic diagram. The effective area of the composite membrane used in the permeation cell is 19.26 cm2. The fluxes of CO2 (NCO2) and CH4 (NCH4) were calculated from the permeate gas flow rate in the downstream. The downstream pressure at the outlet of the permeation cell is 1 atm pressure. The permeability is given below: RCO2 = NCO2/∆pCO2 and RCH4 = NCH4/∆pCH4
Fig. 1. Schematic diagram of the sorption experimental apparatus. 1 sorption chamber, 2 quartz spring, 3 membrane patch, 4 mercury manometer, 5 gas cylinder, 6 vacuometer, 7 vacuum pump, 8 temperature-control system, 9 valve.
316
L.-g. Wu et al. / Desalination 193 (2006) 313–320 Table 1 Effect of the monomer ratio AM/MA on Mv
Fig. 2. Schematic diagram of the permeation experimental apparatus. 1 linear regulator, 2 gas cylinder, 3 thermostatic bath, 4 permeation cell, 5 barometer, 6 valve, 7 flowmeter.
3. Results and discussion 3.1. Synthesis of the AM–MA copolymer The synthesis of the copolymer of AM and MA was carried out using AIBN as an initiator at 80 in a different monomer ratio (AM/MA). The intrinsic viscosities ([η]) and viscosity–average molecular weight (Mv) for the AM–MA copolymer are given in Table 1. The results shown in Table 1 indicate that the molecular weight of the AM–MA copolymer was dependent on the monomer ratio AM/MA, and the molecular weight of the AM–MA copolymer greatly increased with the growth of the monomer ratio AM/MA. For the system of the AM–MA copolymerization, it was found that the AM and MA are easily copolymerized, but the self-polymerization of AM happens more easily than does the copolymerization of AM–MA; and the selfpolymerization of MA happens with difficulty. The polymerization reaction is shown in Fig. 3. 3.2. Structural analysis of the AM–MA copolymer The carbon, hydrogen, and nitrogen content of the AM–MA copolymer was determined by the method of elemental analysis. The polymerization ratio of the monomer (m) and the content of the
Monomer ratio AM/MA, mol/mol
[η], ml/g
Mv ×10!4
5:1 6:1 7:1 8:1
1.72 2.06 2.32 2.79
7.44 9.75 12.60 16.60
Table 2 Effect of the monomer ratio AM/MA on the content of amine group and the polymerization ratio Monomer ratio AM/MA, mol/mol
NH2, wt%
m, mol/mol
5:1 6:1 7:1 8:1
18.08 19.65 20.90 21.47
5.6 9.4 17.6 27.9
amine group in the copolymer were calculated from the carbon and nitrogen content of the copolymer. The results shown in Table 2 indicate that the amine group content in the copolymer slightly increased, the monomer molar ratio in the copolymer greatly increased with the growth of the monomer ratio AM/MA. The effect of the monomer molar ratio on the growth of the amine group content due to the probability of the AM self-polymerization increased with the enlargement of the monomer ratio AM/MA. The structure of the AM–MA copolymer alternates according to the literature data [9] and the IR spectrum (Fig. 4). The characteristic absorption peaks of the IR spectra for the AM and MA unit are 3430, 1700, 1630, 1600, 1350, 560 cm!1, respectively. The broad peak at 3430 cm!1 should be attributed to the N-H stretching vibration in the O=C-NH2 group. The peak at 1630 cm!1 should be attributed to the N-H deformation vibration in the O=C-NH2 group. The peak at 1600 cm!1 should be attributed to the C=O
L.-g. Wu et al. / Desalination 193 (2006) 313–320
317
Fig. 3. Polymerization reaction of AM–MA.
Fig. 4. IR spectra of the AM–MA copolymer. 1 AM/MA = 5:1; 2 AM/MA = 6:1; 3 AM/MA = 7:1; 4 AM/MA = 8:1.
stretching vibration in the O=C-NH2 group. The peak at 560 cm!1 should be attributed to the N-C=O deformation vibration in the O=C-NH2 group. The peak at 1700 cm!1 should be attributed to the C=O stretching vibration in the O=C-C=O group. The peak at 1350 cm!1 should be attributed to the C-O stretching vibration in the O=C-C=O group. 3.3. Sorption behavior Fig. 5 shows the sorption behavior of CO2, CH4 in the AM–MA copolymer at different AM/MA monomer ratios at 25 and 1.05 Mpa. All
CH4 sorption curves can be classified as consistent with the Fickian diffusion model. All CO2 sorption curves can be classified as two-stage sorption. Hence, it appears that the CO2 can react in reverse with the amine group in the AM–MA copolymer. A dual-mode sorption model can be used to describe the sorption behavior of CO2 in the AM–MA copolymer. Fig. 6 shows the effects of the AM/MA monomer ratio on the saturated sorption capacity of CO2 and CH4 in the AM–MA copolymer. From Fig. 6 we find that the copolymer appears high in the saturated sorption capacity of CO2, and low in the saturated sorption capacity of CH4. The
318
L.-g. Wu et al. / Desalination 193 (2006) 313–320
(a)
(b)
Fig. 6. Effects of the AM/MA monomer ratio on the saturated sorption capacity.
AM/MA 5/1. From Fig. 7 it can be seen that the CO2 sorption isotherm of the AM–MA copolymer membrane is in accord with the dual-mode sorption model. The CH4 sorption isotherm of the AM–MA copolymer membrane is consistent with Henry’s law. 3.4. Permeation performance Fig. 5. Sorption behavior of (a) CH4 and (b) CO2 in the AM–MA copolymer.
saturated sorption capacity of CO2 increased slightly with a growth in the monomer ratio of AM/MA, and the saturated sorption capacity of CH4 was kept constant. The effect on the saturated sorption capacity of CO2 increased due to the amine group in the copolymer, which also increased slightly with the enlargment of the monomer ratio AM/MA. Fig. 7 shows the effect of feed gas pressure on CO2 and CH4 saturated sorption capacity of the AM–MA copolymer membrane for a different temperature of feed gas at the monomer ratio
Fig. 8 shows the effect of the AM/MA monomer ratio on the permeation performance of the AM–MA copolymer composite membrane tested with pure CO2 and CH4. From Fig. 8 we find that the composite membrane shows a high CO2 permeation rate and a low CH4 permeation rate. In addition, the CO2 permeation rate increased slightly with the growth of the AM–MA monomer ratio, and the CH4 permeation rate remained constant. These results were consistent with the effect of the AM/MA monomer ratio on the saturated sorption capacity. Fig. 9 shows the effect of feed gas pressure on the permeability of the AM–MA copolymer composite membrane that was tested with pure CO2 and CH4. The CO2 permeation rate decreased
L.-g. Wu et al. / Desalination 193 (2006) 313–320
(a)
319
(b)
Fig. 7. Effect of feed gas pressure on the saturated sorption capacity. (a) CO2. (b) CH4.
(a)
Fig. 8. Effect of the AM/MA monomer ratio on permeability.
with the increase of the feed gas pressure. This is a characteristic of the facilitated transport mechanism. With an increase of the feed gas pressure, CO2 flux (NCO2) increases greatly in the lower pressure region, and the extent of increment decreases gently in the higher pressure region due to the approach of saturation of the limited available carrier. Therefore, the CO2 permeation rate, which equals the flux divided by the pressure difference, decreased with the increase of the feed gas pressure. The CH4
(b)
Fig. 9. Feed gas pressure on permeability. (a) CO2; (b) CH4.
320
L.-g. Wu et al. / Desalination 193 (2006) 313–320
permeation rate remained almost constant with the increase of the feed gas pressure. [3]
4. Conclusions AM–MA copolymeric membranes exhibit high CO2 saturated sorption capacity, high CO2 permeability, and low CH4 saturated sorption capacity, and low CH4 permeability due to CO2 selective sorption preferentially facilitated transport caused by CO2 reacting reversibly with amine groups in the membranes.
[4]
[5]
[6]
Acknowledgements This work was supported by the National Natural Science Foundation of China (No. 20276014) and the State Key Development Program for Basic Research of China (No. 2003CB615700).
[7]
[8]
References [1] L.M. Robeson, The limits of the facilitated transport, J. Membr. Sci., 62 (1991) 165–185. [2] J.H. Kim, B.R. Min, J. Won and Y.S. Kang, Complexation mechanism of olefin with silver ions
[9]
dissolved in a polymer matrix and its effect on facilitated olefin transport, Chem. Eur. J., 8(3) (2002) 650–654. S. Sunderrrajan, B.D. Freeman, C.K. Hall and I. Pinnau, Propane and propylene sorption in solid polymer electrolytes based on poly(ethylene oxide) and silver salts, J. Membr. Sci., 182 (2001) 1–12. Y. Zhang, Z. Wang and S.C. Wang, Facilitated transport of CO2 through synthetic polymeric membrane, Chinese J. Chem. Eng., 10(5) (2002) 570–574. H.L. Chen, L.G. Wu and C.L. Zhu, PVA membrane filled β-cyclodextrin for separation of isomeric xylenes by pervaporation. Chem. Eng. J., 78 (2000) 159–164. H. Matsuyama and K. Matsui, Effects of membrane thickness and membrane preparation condition on facilitated transport of CO2 through ionomer membrane, Sep. Purif. Technol., 17 (1999) 235–241. H. Matsuyama, A. Terada, T. Nakagawara, Y. Kitamura and M. Teramoto, Facilitated transport of CO2 through polyethylenimine/poly(vinyl alcohol) blend membrane, J. Membr. Sci., 163 (1999) 221– 227. W.F. Lee and G.Y. Huang, Poly(sulfobetaine) and corresponding cationic polymers: Synthesis and dilute aqueous solution properties of poly(sulfobetaine)s derived from acrylamide-maleic anhydride copolymer, Polymer, 37(19) (1996) 4389–4395. M.K. Yakovleva, A.P. Sheinker and A.D. Abkin, Vysokomol Soyed (A), 10 (1968) 1946.