Polymer composition has been prepared on the basis of commercial siloxane rubber SKTN-D and tris-(methyldiethoxysiloxy)iron with the reagent ratio 60 : 40 wt %, respectively. Transport parameters of individual hydrocarbons C1–C4 have been shown to be comparable with those for polydimethylsiloxane published previously. Permeability coefficients P of single hydrocarbons C2–C4 increase with an increase in the penetrant feed pressure. It has been demonstrated for n-butane that such increase in P resulted from both a higher diffusion coefficient and solubility coefficient of the penetrant. For mixed gas permeation of C1–C4, an increase in P of methane and a decrease in P of n-butane are observed with increasing the feed pressure. These trends may be explained by the plasticization of the polymer compound and the presence of “rigid” polycyclic structures based on the iron-siloxane component. Similar behavior has been observed for methane and C2+ components with separating quaternary mixture of C1–C4 hydrocarbons. Separation selectivities with respect to methane decreased with increasing feed pressure; however, they become comparable with corresponding results for previously studied substituted polytricyclonones.
Selectively permeable membranes are used in many industrial processes to separate gaseous (or liquid) mixtures. Membrane technology as an energy-efficient process may replace traditional gas purification methods. The importance of membrane gas separation technology has attracted attention in recent decades, both in research area and for industrial applications [1–4].
Membrane gas separation technology is a relatively new and potential alternative method for the extraction of heavy hydrocarbons from natural and associated petroleum gas . To solve these problems, it is advisable to use membranes with so-called thermodynamic selectivity for hydrocarbons . The range of polymers that satisfy this requirement is relatively narrow and includes highly permeable glassy polymers (substituted polyacetylenes [7–11], polynorbornenes [12–17], polybenzodioxane ladder polymer PIM-1 [18, 19]) and highly elastic polymers such as polyorganosiloxanes [20–23]. Among them, polydimethylsiloxane (PDMS) and polyoctylsiloxane [3, 23, 24] were commercialized in the membrane separation field. Moreover, PDMS remains a subject of scientific research, whose purpose is not only a comprehensive study of the process of permeability and diffusion [25–27] but also the improvement of its mechanical characteristics and separation properties [28–31].
The present work is devoted to the study of the gas separation characteristics of the membrane, which is an iron-containing composition based on industrial siloxane rubber SKTN-D and tris-(methyldiethoxysiloxy)iron at the reagent ratio of 60 : 40 wt %, respectively (Fe-PS). The gas transport properties of the test material and other compositions with various multifunctional metallosiloxanes [32, 33] were reported previously ; it was shown that the iron-containing polymer composition has the best gas transport parameters. Due to the high permeability coefficients of n-butane, good selectivity of C4H10/CH4, and improved mechanical properties, Fe-containing polydimethylsiloxanes are promising materials for the manufacture of membranes designed to separate heavy hydrocarbons from natural gas.
Synthesis of tris(methyldiethoxysiloxy)iron. All operations were carried out in an argon medium. A solution of sodium oxymethyldiethoxysilane in dried toluene was added to a suspension of iron(III) chloride in the same solvent. The reaction was carried out at 40°C. Yield was 81%. The synthesis details were reported in .
Preparation of the Fe-containing polymer composition. 60 wt % of SKTN-D and 40 wt % of tris(methyl-diethoxysiloxy)iron were mixed in a toluene solution. The resulting mixture was placed in a Teflon form and kept at room temperature and atmospheric pressure to remove the solvent. The mixture was cured for 1 hour at 70°C and then 2 hours at 200°C. The completion of the process was determined as the constant weight of the resulting film was achieved.
The fabricated polymer composition consists of many three-dimensional polydimethylsiloxane networks with non-aggregated metal atoms (the proposed structure is shown in Fig. 1). The introduction of multifunctional metallosiloxanes into the system is caused by the need for a uniform distribution of iron siloxane structures in the polymer matrix. Metalosiloxanes imparts not only special properties to the material but also significantly improves the mechanical properties of PDMS-based compositions.
The permability P and diffusion D coefficients of individual gases were found using the Danes–Barrer method (time lag method). The measurements were carried out at 22 ± 2°C on a Barotron barometric setup . The submembrane pressure fell in the range 0.5–3.0 bar; the pressure below the membrane did not exceed ~12 Torr (16 mbar). The stationary flow through the polymer film was measured at times (4–6) θ (where θ is the time lag). The calculation of the diffusion coefficient was carried out according to the equation:
where l is the thickness of the film under study.
The mixed-gas permeability was measured by a gas separation unit (Fig. 2), which allows one to determine the gas content in the permeate and retentate using the chromatographic method and to vary the composition and pressure of the mixture supplied. Helium was used to purge the submembrane volume and transfer permeate to a gas chromatograph; the volumetric flow rate of He was 50 mL/min. The total volumetric flow rate of the carrier gas and permeate was measured using a foam flow meter. A constant feed flow rate of 60 cm3/min was maintained during the measurement of the permeability of individual gases.
The following hydrocarbon mixtures were prepared for separation experiments: CH4 : n-C4H10 (90 : 10) and CH4 : C2H6 : C3H8 : n-C4H10 (87 : 6 : 4 : 3). The permeability coefficients of individual gases (CH4, C2H6, C3H8, and n-C4H10) were measured in the above mixtures at the transmembrane pressure falling in the range 1–10 atm.
RESULTS AND DISCUSSION
The permeability coefficients of individual gases of C1–C4 hydrocarbons were determined on the chromatographic and barometric installations. The results of both methods are in good agreement with each other, as can be seen in Table 1. When comparing the hydrocarbon permeability coefficients in the studied Fe-PS composition with the corresponding literature data for PDMS, it was found that Fe-PS has increased values of the permeability of n-butane, while the permeability of the other gases studied is lower (Table 1). The latter may be associated with the formation of a rigid polycyclic structure of iron oxysiloxane, which is much less permeable to gas molecules .
To estimate the effect of hydrocarbon penetrants on the transport parameters of hydrocarbon mixtures, concentration dependences of the permeability of individual hydrocarbons for Fe-PS in the pressure range 0.5–8 atm were obtained (Fig. 3). The permeability coefficient of CH4 is independent of pressure, which is typical for the penetration of light gases in highly elastic polymers . Unlike methane, the permeability of hydrocarbons (C2+) in the pressure range under study increases linearly with increasing pressure (Fig. 3a) and penetrant activity (Fig. 3b) in semi-log coordinates, which is partially due to the plasticization of the membrane material. Plasticization in the presence of condensing gases is caused by an increase in the mobility of macromolecular chains, which leads to an increase in the fraction of free volume and, accordingly, an increase in diffusion coefficients . Easily condensing gases, such as n-C4H10, plasticize the polymer matrix due to their high concentration in the polymer. In addition, the solubility coefficient of n‑C4H10 increases with increasing pressure, which also contributes to an increase in the permeability coefficient .
The pressure dependence of the diffusion coefficient of individual n-butane was studied using the barometric method. As can be seen, an increase in the diffusion coefficient of n-C4H10 with increasing pressure is observed (Fig. 4), which is typical for plasticized polymers [2, 15]. Based on these data, the solubility coefficients of butane in the studied composition Fe-PS were estimated using the expression S = P/D . It can be seen from the obtained data that growth in the n-butane permeability for the polymer composition with increase in pressure is associated with increase in both the diffusion and solubility coefficients. In this case, a gradual linear increase of these parameters is observed with growth of the submembrane pressure to 1.1 atm.
Separation of Two-Component Mixture
The study of gas separation characteristics of the material using model mixtures of hydrocarbons that are close to real is a highly desirable research stage. Table 2 shows data on the permeability of a binary methane-butane mixture for some polymers. It can be seen from the data presented that the study of the permeability of a binary hydrocarbon mixture for the studied composition Fe-PS should also be of interest due to the high values of the permeability coefficient of n-butane (14 900 Barrer) and ideal selectivity α[C4Н10/CН4] = 22.
During the separation of the two-component mixture CH4–n-C4H10 (90/10), we studied the effect of the volumetric rate and pressure of the feed stream on transport and separation parameters of the composition Fe-PS. The results of these studies are shown in Fig. 5. It is seen that there is an increase in the methane permeability coefficient with increasing pressure of the feed stream of the mixture. Such behavior for methane in two-component mixtures was previously described in [27, 39]. The high concentration of n‑C4H10 in the polymer (its solubility in PDMS at 1 atm can reach 56 cm3 (STP) cm–3 (polymer) atm–1 ) makes the polymer matrix more similar to the medium of liquid butane, in which methane has a high solubility . Thus, the presence of n-butane in the polymer creates a more favorable environment for the sorption and transport of methane, which leads to an increase in the solubility and, consequently, the permeability of CH4.
The change in the volumetric rate of the feed stream of the binary hydrocarbon mixture has no significant effect on the apparent coefficient of methane permeability in the mixture (Fig. 5).
For n-butane, on the contrary, the apparent permeability coefficient decreases with increasing pressure of the feed stream of the mixture. This is not consistent with earlier studies such as [27, 38], where the presence of methane results in no changes in the permeability of butane. It is likely that the difference between the observed data and the literature is associated with structural features of the Fe-PS composition. Namely, in the “rigid” polycyclic framework, the mobility of the metallosiloxane chains is reduced, which contributes to the appearance of “frozen” free volume elements; this is similar to the appearance of metastable vacancies in a glassy polymer because of the low mobility of fragments of macromolecular chains . In these free volume elements, “competitive” sorption of penetrant molecules occurs, which leads to a decrease in the solubility of n-butane and, consequently, to a decrease in its permeability. However, confirmation of this assumption requires additional studies of the relaxation properties, as well as sorption and diffusion of the mixture of hydrocarbons C1 and C4 for different iron-siloxane compositions. It can be assumed that the observed phenomenon is similar to the behavior described previously by Genduzo et al.  upon separation of the CO2–CH4 mixture. With an increase in the penetrant pressure, the authors first observed a decrease in the coefficient of permeability of CO2, and then its increase, which they explained by the effect of competitive sorption.
In Fig. 5, one can trace the effect of the feed stream velocity on the apparent permeability of n‑butane resulting from the appearance of concentration polarization on the membrane surface. It was previously reported [41, 42] that during the separation of gas mixtures, where the more permeable component has a penetration rate of more than 100 GPU [1 GPU = 10–6 cm3 (STP) cm–2 s–1 (cm Hg)–1] and selectivity with respect to the other component is more than two, concentration polarization can significantly affect the permeability coefficients. In the present work, the penetration rate of butane at 1 atm is approximately 500 GPU, and the ideal selectivity of C4H10/CH4 reaches 22. Thus, the accumulation of a less permeable component (methane) and depletion of a more permeable component (butane) in the boundary layer are possible.
The effect of concentration polarization on the apparent permeability coefficient of n-C4H10 in the mixture is more pronounced at the minimum studied feed flow rate of 40 mL/min, at which the butane extraction rate can reach more than 30% at a mixture pressure of 8 atm (Fig. 6). However, for a feed flow rate of the mixture equal to 160 mL/min, this effect is almost leveled as can be seen from Fig. 5b.
A comparison of the permeability coefficients of individual n-butane and its mixture can be seen in Fig. 7. It is clear that a regular increase in the permeability coefficients is found for the individual component, while their decrease is observed (Fig. 7a) for the binary mixture. This behavior significantly changes the concentration dependence of the selectivity of the C4H10/CH4 separation. Thus, the ideal selectivity of butane-methane separation increases with increasing pressure of the mixture, while for the binary mixture, a gradual decrease in selectivity is observed. Thus, the lower the content (partial pressure) of butane in the binary mixture, the more efficient the separation of the methane–butane mixture.
Separation of Four-Component Mixture
The gas-separation characteristics of the Fe-PS composition were studied using the four-component mixture CH4 : C2H6 : C3H8 : n-C4H10 with a component ratio of 87 : 6 : 4: 3. The experiment was carried out in two modes: (a) using He as a purge gas which was supplied at a rate of 50 mL/min; and (b) in the absence of any carrier gas, when the permeate flow rate was determined only by the flow rate of the gas penetrating through the membrane.
In the case of permeate flushing with helium in the submembrane space, the permeability coefficient of hydrocarbons C2+ decreased with increasing mixture pressure, while methane permeability increased (Fig. 8a). Similar to the separation of the two-component mixture, plasticization of the polymer composition occurs due to the presence of highly soluble hydrocarbons, which increases the permeability of methane. A decrease in the permeability coefficients of the remaining components with increasing pressure of the feed stream can be explained by the competitive sorption phenomenon.
A different picture is observed for permeability measurements without using the purge gas. An increase in the permeability coefficients of each component in the mixture with an increase in the pressure of the feed stream is observed (Fig. 8b). Despite this tendency, the permeability values of each component are lower than in the experiment using flushing by the carrier gas. This trend is understandable if one takes into consideration of the differences in the driving force (concentration gradient in the membrane) in both cases.
As can be seen from Fig. 9, the separation selectivities of the above gas pairs are similar for the presence and absence of helium flushing and reach maximum value at a feed pressure of the four-component mixture equal to 2 atm. A gradual decrease in selectivities for the given polymer composition with increasing pressure (Fig. 9) is associated with the increase in the methane permeability coefficient (Fig. 8a). Thus, on the basis of the results obtained, it can be concluded that the maximum separation selectivities of the four-component hydrocarbon mixture C1–C4 can be achieved using a feed stream pressure close to the atmospheric pressure (1–2 atm).
The obtained hydrocarbon separation selectivities for the iron-containing composition studied turned out to be close to or higher than the results obtained previously in the separation of four-component mixtures for highly permeable polymers based on substituted tricyclononenes (P(C2H6)/P(CH4) = 1.2–1.6, P(C3H8)/P(CH4) = 2.1–3.8, and P(C4H10)/P(CH4) = 5.3–9.3) .
The gas separation parameters of a polymer composition based on PDMS (SKTN-D) and metallosiloxane, namely tris-(methyldiethoxysiloxy)iron, have been measured using two- and four-component C1–C4 hydrocarbon mixtures. The methane permeability increases while the permeabilities of the other hydrocarbon components decrease with the increase of feed stream pressure for the both hydrocarbon mixtures. This can be explained by plasticization of the polymer matrix and the appearance of the competitive sorption effect. The maximum selectivity values (P(C4H10)/P(CH4) = 10) are achieved during the experiment with a feed pressure of 2 atm.
R. W. Baker, Membrane Technology and Applications, 2nd ed. (Wiley & Sons Inc., Chichester, England, 2004).
S. Matteucci, Y. Yampolskii, B. D. Freeman, and I. Pinnau, in Materials Science of Membranes, Ed. by Y. Yampolskii, I. Pinnau, and B. D. Freeman (John Wiley & Sons, Ltd., Chichester, 2006).
P. Bernardo, E. Drioli, and G. Golemme, Ind. Eng. Chem. Res. 48, 4638 (2009).
A. F. Ismail, K. C. Khulbe, and T. Matsuura, Gas Separation Membranes: Polymeric and Inorganic (Springer Int. Publishing, Switzerland, 2015).
J. Schultz and K.-V. Peinemann, J. Membr. Sci. 110, 37 (1996).
Y. Yampolskii, L. Starannikova, N. Belov, M. Bermeshev, M. Gringolts, and E. Finkelshtein, J. Membr. Sci. 453, 532 (2014).
K. Nagai, L. G. Toy, B. D. Freeman, M. Teraguchi, T. Masuda, and I. Pinnau, J. Polym. Sci., Part B: Polym. Phys. 38, 1474 (2000).
K. Nagai, L. G. Toy, B. D. Freeman, M. Teraguchi, G. Kwak, T. Masuda, and I. Pinnau, J. Polym. Sci., Part B: Polym. Phys. 40, 2228 (2002).
L. G. Toy, K. Nagai, B. D. Freeman, I. Pinnau, Z. He, T. Masuda, M. Teraguchi, and Y. P. Yampolskii, Macromolecules 33, 2516 (2000).
R. D. Raharjo, H. J. Lee, B. D. Freeman, T. Sakaguchi, and T. Masuda, Polymer 46, 6316 (2005).
D. A. Syrtsova, O. B. Borisevich, O. A. Shkrebko, V. V. Teplyakov, D. D. Grinshpan, V. S. Khotimskii, and D. Roizard, Sep. Purif. Technol. 57, 435 (2007).
Y. Grinevich, L. Starannikova, Y. Yampolskii, M. Gringolts, and E. Finkelshtein, J. Membr. Sci. 378, 250 (2011).
Y. Grinevich, L. Starannikova, Y. Yampol’skii, M. Gringol’ts, and E. Finkel’shtein, Pet. Chem. 51, 585 (2011).
Y. V. Grinevich, L. E. Starannikova, Y. P. Yampolskii, and M. V. Bermeshev, Polym. Sci., Ser. A 55, 43 (2013).
N. Belov, Y. Nizhegorodova, M. Bermeshev, and Y. Yampolskii, J. Membr. Sci. 483, 136 (2015).
N. Belov, R. Nikiforov, L. Starannikova, K. R. Gmernicki, C. R. Maroon, and B. K. Long, Eur. Polym. J. 93, 602 (2017).
B. J. Sundell, J. A. Lawrence III, D. J. Harrigan, J. T. Vaughn, T. S. Pilyugina, and D. R. Smith, RSC Adv. 6, 51619 (2016).
S. Thomas, I. Pinnau, N. Du, and M. D. Guiver, J. Membr. Sci. 333, 125 (2009).
P. Li, T. S. Chung, and D. R. Paul, J. Membr. Sci. 432, 50 (2013).
B. S. R. Reddy and U. Senthilkumar, J. Sci. Ind. Res. 62, 666 (2003).
E. A. Grushevenko, I. L. Borisov, D. S. Bakhtin, S. A. Legkov, G. N. Bondarenko, and A. V. Volkov, Petr. Chem. 57, 334 (2017).
E. A. Grushevenko, I. L. Borisov, D. S. Bakhtin, G. N. Bondarenko, I. S. Levin, and A. V. Volkov, React. Funct. Polym. 134, 156 (2019).
H. Mushardt, M. Müller, S. Shishatskiy, J. Wind, and T. Brinkmann, Membranes 6, 16 (2016).
R. W. Baker, Membranes for Vapor/Gas Separation (Membrane Technology and Research, Menlo Park, 2006).
T. C. Merkel, V. I. Bondar, K. Nagai, B. D. Freeman, and I. Pinnau, J. Polym. Sci., Part B: Pol. Phys. 38, 415 (2000).
I. Pinnau and Z. He, J. Membr. Sci. 244, 227 (2004).
R. D. Raharjo, B. D. Freeman, D. R. Paul, G. C. Sarti, and E. S. Sanders, J. Membr. Sci. 306, 75 (2007).
M. Rezakazemi, A. Vatani, and T. Mohammadi, J. Nat. Gas Sci. Eng. 30, 10 (2016).
M. S. Suleman, K. K. Lau, and Y. F. Yeong, J. Appl. Polym. Sci. 135, 45650 (2018).
J. Yang, M. M. Vaidya, V. V. R. Tammana, and D. Harrigan, U.S. Patent No. 10293301 (2019).
M. Sadrzadeh, K. Shahidi, and T. Mohammadi, J. Appl. Polym. Sci. 117, 33 (2010).
A. M. Muzafarov, N. A. Tebeneva, E. A. Rebrov, N. G. Vasilenko, M. I. Buzin, and N. V. Nikolaeva, RF Patent No. 2296767 (2007).
A. M. Muzafarov, N. A. Tebeneva, E. A. Rebrov, N. G. Vasilenko, M. I. Buzin, and N. V. Nikolaeva, RF Patent No. 2293746 (2007).
N. A. Belov, A. N. Tarasenkov, N. A. Tebeneva, G. N. Vasilenko, and G. A. Shandryuk, Yu. P. Yampolskii, and A. M. Muzafarov, Polym. Sci., Ser. B 60, 405 (2018).
N. A. Belov, R. Yu. Nikiforov, M. V. Bermeshev, Yu. P. Yampol’skii, and E. Sh. Finkel’shtein, Pet. Chem. 57, 923 (2017).
I. Pinnau and L. G. Toy, J. Membr. Sci. 116, 199 (1996).
W. J. Koros and M. W. Hellums, in Encyclopedia of Polymer Science and Technology, Ed. by J. I. Kroschwitz (Wiley, NewYork, 1990), pp. 724–802.
R. D. Raharjo, B. D. Freeman, and E. S. Sanders, J. Membr. Sci. 292, 45 (2007).
G. Genduso, B. S. Ghanem, and I. Pinnau, Membranes 9, 10 (2019).
A. P. Safronov and L. V. Adamova, Polymer 43, 2653 (2002).
K. Haraya, T. Hakuta, H. Yoshitome, and S. Kimura, Sep. Sci. Technol. 22, 1425 (1987).
G. He, Y. Mi, P. L. Yue, and G. Chen, J. Membr. Sci. 153, 243 (1999).
B. N. Ershov, E. E. Khoruzhii, Z. K. Olenina, V. K. Belyakov, and F. A. Makhmutov, Plast. Massy 9, 23 (1983).
This work was carried out within the State Program of the Topchiev Institute of Petrochemical Synthesis, Russian Academy of Sciences. The manufacture of PDMS-metallosiloxane compositions was performed within the State Program of the Enikolopov Institute of Synthetic Polymeric Materials, Russian Academy of Sciences. The synthesis of functional metallosiloxane oligomers was supported by the Russian Foundation for Basic Research, project no. 16-29-10781.
Translated by V. Avdeeva
About this article
Cite this article
Bezgin, D.A., Belov, N.A., Nikiforov, R.Y. et al. Separation of C1–C4 Hydrocarbon Mixtures Using Fe-Containing Siloxane Composition. Membr. Membr. Technol. 2, 27–34 (2020). https://doi.org/10.1134/S2517751620010035
- iron siloxane
- mixed-gas permeation