(180ae) Mass Transport Properties of Gases and Vapors in Matrimid Polyimide

Luca Ansaloni, Marco Giacinti Baschetti, Matteo Minelli, Giulio Cesare Sarti

Dipartimento di Ingegneria Chimica, Mineraria e delle Tecnologie Ambientali

Centro Interdipartimentale di Ricerca Industriale Meccanica Avanzata e Materiali

Alma Mater Studiorum - Università di Bologna, via Terracini 28, 40131 Bologna, Italy

Membrane-based separation process are gaining interest in many industrial application in view of the low energy consumption and the high selectivity potentially achievable. In particular, a large portion of gas separation studies has been focused on glassy polymeric membranes which are able to combine good selectivity behaviour and high fluxes, essential prerequisites to compete with other separation methods. However, glassy polymers are inherently non equilibrium materials, which on a long time scale tend to relax towards the configuration of true thermodynamic equilibrium.

Their physical and chemical properties, therefore, are particularly influenced by polymer history [1-6] and physical aging [7,8]. In addition, the presence of minor components on the feed flow, such as impurities or water vapour, can strongly affect the transport properties, altering the separation performances [9,10].

Aromatic polyimides are attractive materials for gas separation due to high T_g and good mechanical and chemical resistance [11]. They show interesting selectivity for CO₂ with respect to other gases such as N₂ and CH₄, while their main weakness is due to their glassy nature, which leads to plasticization phenomena at high fugacity of gases and vapors [12,13], and to aging phenomena that decrease the performances during time [7,8,14].

Matrimid 5218 polyimide, which is probably the most known among the commercially available polyimides, has been investigated in this work, focusing the attention on different thermal pre-treatments, physical aging and the effect of water vapor on the transport properties. The study of permeation of humid gases allows the characterization of a wide range of possible operative conditions of real applications, since water vapor is normally present in the feed stream of several industrial processes [15].

Free standing Matrimid membranes, with thickness of 30-50 µm, have been cast from a 1% wt. solutions of polymer in CH₂Cl₂. Samples have been treated following different thermal histories, consisting in keeping them under vacuum at different temperatures (50, 100, 150 and 200°C) for 24 hours. All different protocols ensured complete solvent evaporation, but led to different glassy states of the final specimen. The pure gas (CO₂, CH₄, H₂), as well as humid gas permeabilities have been investigated with a barometric apparatus, monitoring the time evolution of pressure in the downstream volume, previously calibrated [16,17]. The upstream pressure has been kept at a constant value (~2 bar), far below the CO₂ plasticization pressure reported in literature for Matrimid [12,13]. Three different operative temperatures have been investigated, 35, 45 and 55°C, to estimate the activation energy of permeation.

Initial tests show that the permeability value for both CH₄ and CO₂ decreases with the increase of temperature of thermal pretreatment with a corresponding enhancement of ideal selectivity, as usually observed in systems that follow the solubility-diffusivity model.[ 18,19]

Furthermore, several permeability tests have been performed at different aging times, covering a period of more than 3000 hours. Results indicate an appreciable decrease of permeability as consequence of aging, for all samples, and for both gases. The effect of aging is more evident in the samples pre-treated at the lower temperatures and is less pronounced with increasing the temperature of thermal pre-treatment.

When water vapor was added to the gas feed, the permeability displayed a marked decline, as often reported for other low water sorbent polymers [9,20], reaching values of about 50% of those observed for dry gases. This decrement, likely related to competitive sorption behavior between water and the permeating gas, was similar for all the gases investigated and under all the studied conditions the selectivity value remained substantially constant at different water amount in the gas stream. This is in contrast to what presented by Chen et al. [10], who reported a more enhanced decrease of CO₂ permeability and a consequent reduction of CO₂/CH₄ selectivity.

Transport properties and aging rates detected at different temperatures of the thermal pretreatment are likely related to changes in density, and consequently in the free volume, of the polymer [14,21-23]. Density measurements were performed with buoyancy method weighting the specimens in air and in liquid n-dodecane [24]. Data have been obtained for all the samples and at different aging times in order to correlate the fractional free volume available in the polymeric matrix with the transport properties [25]. Due to the chains compaction, a higher rigidity level is achieved, leading to an increase of selectivity values. As a result, the samples pretreated at lower temperature (50 and 100°C) display initially a faster aging rate, but after a relative short period of time, they tend to behave similarly to the samples pretreated at higher temperatures. Therefore, a physical aging model has been implemented, in order to identify the characteristic aging time, which describe the behavior of the different samples [22,23].

To better understand the H₂O behavior in Matrimid, pure water permeation and water vapor sorption experiments were carried out at 3 different temperatures (25, 35 and 45°C). The sorption isotherms show a linear behavior at low activity whereas at higher R.H. an upward curvature is apparent. Experimental permeabilities obtained either from steady permeation or from the solubility and diffusivity data from transient sorption are consistent with each other, in accordance to the general scheme of solution-diffusion mechanism. The water uptake behaves similarly with respect to water activity at the three temperatures investigated.

The monotonic decrease of gas permeability under humid conditions points out that the competitive sorption predominates over any plasticization effects, also in view of the limited amount of water that can be absorbed in the polymer.

Finally, permeability results are analyzed by means of specific models in order to better understand the different aspects experimentally investigated. Specifically, the behavior of solubility of gases and water vapor in Matrimid are treated with a nonequilibrium thermodynamic model (NELF) [26,27] able to describe sorption capacity of glassy polymers. The model considers the polymer density (and consequently its free volume) as a measure of the deviation from the equilibrium state [28,29].

Diffusivity of low molecular weight species in Matrimid are then analyzed with a free-volume based model, able to represent the different configurations of the polymer obtained after different pretreatments or aging phenomena.

References:

1. Kawakami H., Mikawa M., Nagaoka S., J. Membr. Sci. 118 (1996) 223-230
2. Galvani F., Ruvolo-Filho A., Pessan L. A., J. Macromol. Sci., B: Polym. Phys. 46 (2007) 931-948
3. Wonders A.J., Paul D.R., J. Membr. Sci. 5 (1979) 63-75
4. Jordan S.M., Koros W.J., Fleming G.K., J. Membr. Sci. 30 (1987)  191-212
5. Hachisuka H., Takizawa H., Tsujita Y., Takizawa A. Kinoshita T., Polymer 32 (1991) 2382-2386
6. Chan A.H., Paul D.R., J. Appl. Polym. Sci. 24 (1979) 1539-1550
7. Huang Y., Paul D.R., Physical, Polymer 45 (2004) 8377-8393
8. Huang Y., Wang X., Paul D.R., J. Membr. Sci. 277 (2006) 219-229
9. Moore T.T., Vo T., Mahajan R., Kulkarni S., Hasse D., Koros W.J., J. Appl. Polym. Sci. 90 (2003) 1574-1580
10. Chen Q.G., Scholes C.A., Qiao G.G., Kentish S.E., J. Membr. Sci. 379 (2011) 479-487
11. White L. S., Blinka T.A., Kloczewski H.A., Wang I., J. Membr. Sci. 103 (1995) 73-82
12. Bos A., Pünt I.G.M., M.Wessling, Strathmann H., Sep. Purif. Tech. 14 (1998) 27-39
13. Tin P. S., Liu Y., Wang R., Liu S. L., Pramoda K. P., J. Membr. Sci. 225 (2003) 77-90
14. Rowe B.R., Freeman B.D., Paul D.R., Polymer 50 (2009) 5565-5575
15. Scholes C.A., Chen Q.G., Tao W.X., Bacus J., Anderson C., Stevens G.W., Kentish S.E., Energy Procedia 4 (2011) 681-687.
16. Minelli M., De Angelis M.G., Doghieri F., Marini M., Toselli M., Pilati F., Eur. Polym. J. 44 (2008) 2581-2588
17. Minelli M., Giacinti Baschetti M., Doghieri F., Ankerfors M., Lindström T., Siró I., Plackett D., J. Membr. Sci. 358 (2010) 67-75
18. Robeson L.M., J. Membr. Sci.  62 (1991) 165-185
19. Robeson L.M., J. Membr. Sci. 320 (2008) 390-400
20. Scholes C.A., Tao W.X., Stevens G.W., Kentish S.E., J. Appl. Polym. Sci. 117 (2010) 2284-2289
21. Volynskii L., Efimov V., Bakeev N.F., Polym. Sci. Ser. C. 49 (2007) 301-320
22. Struik L.C.E., Elsevier, Amsterdam, 1978
23. Struik L.C.E., Polymer 28 (1987) 1869-1875
24. Shishatskiy S., Nistor C., Popa M., Pereira Nunes S., Peinemann K.V., Adv. Eng. Mater. 8 (2006) 390-397
25. Park J.Y., Paul D.R.,  J. Membr. Sci. 125 (1997) 23-39
26. Doghieri F., Sarti G.C., Macromolecules 29 (1996) 7885-7896
27. Sarti G.C., Doghieri F., Chem. Eng. Sci. 53 (1998), 3435-3447
28. Giacinti Baschetti M., Doghieri F., Sarti G.C., Ind. Eng. Chem. Res. 40 (2001) 3027-3037
29. Minelli M., Campagnoli S., De Angelis M.G., Doghieri F., and Sarti G.C., Macromolecules 44 (2011) 4852-4862