Silica Modified Membrane for Carbon Dioxide Separation from
Natural Gas
Mohammed Nasir Kajama
1*, Professor Yilmaz Yildirim
2, Usman Habu Taura
3,
Alhaji Shehu Grema
4,
and Shehu Abdulrahman
51,5Department of Mechanical Engineering, University of Maiduguri, Nigeria
2Department of Environmental Engineering, Bulent Ecevit University, Zonguldak, Turkey
3,4Department of Chemical Engineering, University of Maiduguri, Nigeria
a[email protected]*, b[email protected], c[email protected],
d[email protected], e[email protected]
Keywords: Dip-coating technique; silica membrane; single gas permeation and selectivity; carbon
dioxide separation.
Abstract. A dip-coating technique was applied to prepare a selective membrane on a commercial
ceramic mesoporous support. Single gas components used for permeance and selectivity were CH4,
CO2, H2, He, N2, and O2 (BOC UK) with at least 99.999 (% v/v) purity. The permeances and
selectivities were obtained at room temperature and transmembrane pressure differences between 0.05 up to 5.0 barg. Gas permeation experiments showed the permeance of CO2 to be strongly
influenced by surface diffusion mechanism. Single gas experiment showed linear flow dependence on the inverse square root of molecular weight at room temperature and 1.0 barg. The single gas selectivities were found to be higher than the ideal Knudsen separation mechanism. The highest CO2/CH4 selectivity value of 24.07 was obtained at 0.7 barg and room temperature.
Introduction
The prevention of environmental smog from industrial sources which occurs through fossil fuel combustion is now receiving considerable attention worldwide. In the past 20 years, the international community has agreed to cut down greenhouse gases under the Kyoto protocol in 1997 [1]. Also, at the United Nations Climate Change Conference Doha (2012), a resolution was adopted to extend the reduction in carbon emissions by 2020 under a second commitment period of the Kyoto Protocol [2]. Environmentalists have also brought in emergency regulations for the reduction of the flare activities that are responsible for a significant fraction of the emissions of greenhouse gases such as carbon dioxide (CO2) and methane (CH4) [3]. Elimination of gas flaring can be
achieved by increasing the amount of recovery through separation and collection processes, disposal processes, capture processes, and utilization processes [1]. Therefore, it is imperative to build up new effective technologies to mitigate these emissions. Inorganic membranes applications are technically significant in environmental issues like separation, catalytic reactions among others [1, 4-7]. Inorganic membranes for CO2 removal can be applied in-situ without any phase change.
Catalytic combustion/oxidation is also being researched as a substitute process for the removal of these pollutants since it is flexible and it requires low energy compared to thermal oxidation [8].
According to the International Union of Pure and Applied Chemistry (IUPAC), porous membranes are classified as; Micropores 0.5 - 2 nm where separation is based on molecular sieving mechanism, mesopores 2 - 50 nm where Knudsen diffusion mechanism is the dominant flow but multilayer flow and/or capillary condensation and viscous flow can also take place, and macropores > 50 nm where there is no separation and the flow mechanism is basically influenced by viscous flow [9-11].
Gas transport mechanisms through porous membranes are influenced by viscous flow, Knudsen diffusion, surface diffusion, multi-layer diffusion, capillary condensation, molecular sieving and solution-diffusion [12-14]. Although, the main transport mechanisms of gas separation through mesoporous alumina membranes are Knudsen diffusion and surface diffusion.
All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of Trans Tech Publications, www.scientific.net. (#108764492-22/06/18,15:14:22)
Viscous flow also known as Poiseuille flow occurs if the mean free path (average distance travelled by a, gas molecule from one collision to the other) is much smaller than the pore diameter. The flow characteristics are determined primarily by collisions among the molecules and no separation is obtained between the different gaseous components. In gas membrane separation processes, viscous flow mechanism is an unwanted transport mode because it is non-selective. In viscous flow, the flux can be written as [15].
) 8 ( 2 RTL P P r P av v πµ ε ∆ = (1) where; 2 2 1 P P Pav + = (2) Where ε is the porosity of the membrane, r is the mean pore radius (m), Pav is the average
pressure (Pa), ΔP (P1 – P2) is the pressure difference (Pa) between the high pressure side and low
pressure side, μ is the viscosity (Pa-s), R gas constant (8.314 J.K-1.mol-1) and T is the permeation
temperature (K) and L is the thickness of the membrane (m).
Knudsen diffusion occurs if the mean free path of the gas molecule is larger than the pore diameter. If the collisions among the permeating molecules and the pore wall of the membrane are the main transport mechanism, the separation is based on molecular weight difference [12, 13, 15]. Thus, Knudsen permeance states that the permeation flux is proportional to the inverse square root of the gas molecular weights. The ideal separation factor is defined as the ratio of the permeability of two different permeating gases and is inversely proportional to the square root of the molecular weight ratio, if the separation is mainly driven by Knudsen diffusion which is written as [15];
5 . 0 ) 2 ( 3 8 RTM L P r P p kn τ π ε ∆ = (3) Where r is the mean pore radius (m), P is the pressure (Pa), τ is the tortuosity, L is the thickness of the membrane (m), π = 22/7, R gas constant (8.314 J.K-1.mol-1) and T is the permeation
temperature (K) and M is the molecular weight of gas (g/mol).
Surface diffusion occurs if the diffusing molecules adsorbed on the pore walls of the membrane and migrates along the pore surface. The rates of surface diffusion are determined by the surface diffusion coefficient and adsorption equilibrium [16]. In order to enhance surface diffusion mechanism, silica substance which is selective to carbon dioxide adsorption is frequently applied in modifying macro- or mesopore membranes without cracks or pinholes [1].
Multi-layer diffusion occurs if the flow of gas molecules is adsorbed in the membrane at a different number of layers. Gas mixture permeates through the pores of the membrane at a given pressure and temperature.
Molecular sieving is used to separate gas molecules which differ in kinetic diameter. It enables the permeation of gases which have lower kinetic diameter to pass through the membrane “sieve” than the larger ones [13]. Zeolite membranes possess pore sizes that are of the same size of the gas molecules and can result in very high separation factors [17].
Solution-diffusion separation relies on the physical-chemical interaction of gases and the dense membrane that determine the amount of gas which accumulates in the membrane matrix [14].
There are various methods which are used to improve membrane selectivity, these are sol-gel technique and chemical vapour deposition (CVD) [15]. Sol-gel technique is considered as the most significant technique for the production of meso- and microporous membranes. It was first used for membranes by Leenaars [18] to develop ceramic ultrafiltration membranes [18, 19]. Dip coating is the most general coating technique utilized with sol-gel chemistry [19]. On the other hand, CVD is
a method which allows coating of the membrane by depositing the ceramic layer by chemical reactions at high temperature [19].
There are three parameters which determine the performance of an efficient membrane. These parameters are the permeability, selectivity and the service life of the membrane. It can be elaborated that the higher the permeability, the lower the membrane area is required. Also, the higher the selectivity, the more efficient the process, the lower the driving force required to attain a separation and therefore the lesser the operating cost of the separation system [20].
Permeability is termed as the flux of a specific gas component through the membrane per unit
of area at a given pressure gradient taking into account the membrane thickness [14, 16].
Pe = qL/A∆P (4)
where Pe is the Permeability [mol-m/(m2.s.Pa)], q is the molar flow (mol/sec), A is the surface area
of the membrane (m2), ∆P is the pressure difference (Pa) across the membrane and L the thickness
of the membrane (m).
Selectivity is the ability of the membrane to separate (select) the required component from the
feed mixture [14]. Selectivity is defined as the ratio of the pure component permeabilities (Py and
Pz) for single gases. It can be written as;
αy,z = Py/Pz (5)
where Py is the permeability of y component (mol-m/m-2 s-1 Pa-1), and Pz is the permeability of z
component (mol-m/m-2 s-1 Pa-1).
Surface diffusion has been named as an important mechanism in a number of studies on silica membranes for the selective separation of CO2.Research, then, has focused primarily on separation
of CO2 from N2 and CH4, the surface diffusion mechanism not being considered strong enough for
the more difficult separation from H2. Way and Roberts [21] used microporous hollow fiber silica
membranes for gas separations. They proposed that both the surface diffusion and the molecular sieving mechanism contribute to the permeation properties of the silica hollow fiber membranes. High permselectivities were observed for CO2/N2 of 28 at 40 °C.
In this study, a commercial ceramic support with 30 nm nominal pore diameter was used as the support. Membrane was synthesized via dip-coating method. The membrane permeability was investigated with single gas permeation experiments at room temperature.
Experimental
Membrane preparation and characterization
Commercial porous alumina supports (symmetric membranes) of tubular configuration supplied by CTI (France) consisted of 77% alumina + 23% TiO2 and an average pore diameter of 30 nm
were employed for this experiment. The alumina support had an internal and outer 7 and 10 mm diameter respectively. It possesses a permeable length of 348 mm and 45% porosity. Pressures between 0.05 up to 5.0 bar and temperatures of 25 0C (298 K) were applied. The membrane design
is shown in Fig. 1. The inner, outer and cross-sectional surface morphologies of the support and membrane were analyzed by scanning electron microscopy (SEM) (Zeiss EVO LS10). The support was found to be defect free.
Figure 1: Tubular membrane design.
The dip-coating solution was prepared by mixing 50 ml of silicon elastomer (Sylgard®) and nine
parts of isopentane contained in a glass tube to obtain a clear and colourless solution. A curing agent (Sylgard®) equivalent to one-tenth of the elastomer was added and the resulting solution was
mixed at room temperature. The solution was then allowed to age for 30 minutes after which the ceramic support was immersed for 30 minutes. The membrane was then oven dried at 65 0C for 24
hours [7, 22] to form an ultra-thin layer on the support. The same procedure was repeated for subsequent coatings. Up to five dip-coated outside the surface of the membranes were prepared and evaluated at room temperature in this experiment.
The experimental set-up consisted of a membrane reactor, gas delivery system for pure gases and gas mixtures, a permeate and retentate exit, a flow meter and thermocouples fixed on the reactor as shown in Fig. 2. However, all connections were being tested for leaks by means of a soap solution using helium. The gas species used in the experiments consisted of single gases CH4, CO2, H2, He,
N2 and O2 (BOC UK) with at least 99.999 (% v/v) purity. In a typical experiment, the gas was
passed into the shell-side and permeated through the coated membrane at different pressures. Gas leak was maintained using graphite seals at each end of the reactor. The permeate was connected to the flowmeter to measure the flow rates. The digital mass flow meter was corrected (set) for each gas.
CO2 H2 CH4
Results and Discussion
Fig. 3 shows the SEM images of the outer surface of the silica membrane. It can be obviously seen that some amount of silica has been adsorbed by the membrane during the dip-coating process forming a dense layer on the outside surface of the support.
Figure 3: SEM image of the silica membrane outer surface. Effect of feed pressure on gas permeation at room temperature
Fig. 4 shows N2, CH4, H2 and CO2 single gas permeance against feed pressure across unmodified
γ-Al2O3 membrane at pressures between 0.5 up to 1.0 bar and room temperature. As can be seen on
Fig. 4, permeances are depended on feed pressure. Theses permeances occurred based on their respected gas molecular weight. On the one hand, lighter gas e.g. H2 with 2 g/mol recorded higher
permeance. H2 permeance of nearly 7x10-4 mol / (m2.s.pa) at 1.0 bar was obtained. On the other
hand, CO2 with higher molecular weight of 44 g/mol obtained a permeance of nearly
3x10-4 mol / (m2.s.pa) at the same condition. This could be attributed to the influence of Knudsen
diffusion gas transport mechanism, which states that gas transport can occur on a porous media (especially mesoporous with pore diameter between 2 to 50 nm) based on their respected molecular weight, and this is confirmed in this experiment by using a 30 nm porous membrane. In general, the permeances of each gas compared to their respected molecular weights occurred as follows; H2>CH4>N2>CO2 = 2>16>28>44 (g/mol). The obtained results corroborates with the literature [23].
Therefore, modification of the alumina support is required in order to allow CO2
separation/transport across the coated membrane.
Silica crystal
Figure 4: Single gas permeance against feed pressure across unmodified alumina support at 298 K.
Figure 5: Single gas permeance against feed pressure across silica membrane at 298 K. Contrary to Fig. 4, Fig. 5 shows N2, CH4, H2 and CO2 single gas permeance against feed pressure
across silica γ-Al2O3 membrane at pressures between 0.5 up to 1.0 bar and room temperature. It can
be seen from Fig. 5 that, CO2 despite its higher molecular weight (44 g/mol), as this is not the case
of molecular weight compared to Fig. 4 on the same condition, obtained the higher permeance where Knudsen diffusion gas transport mechanism is negligible. Ohwoka, Ogbuke & Gobina, (2012) [24] claimed a successful CO2 permeation on silica membrane at room temperature and a
feed pressure between 0.1 up to 1.0 bar. Their CO2 permeation increases exponentially with
trans-membrane pressure. In this experiment, the higher CO2 transport achieved through the silica
membrane is due to adhesion of silica on the alumina support via the repeat up to five times dip-coating method which is attributed to surface diffusion gas transport mechanism which corroborates the literature [24].
Fig. 6 shows the influence gas flow rate across silica membrane after fifth dip-coating against feed pressure at 298 K. It can be seen that nitrogen and methane recorded zero flow between 0.05 to
0.7 barg. However, permeation occurred at 1.0 to 5.0 barg but the recorded flow rate was lower than carbon dioxide and hydrogen. This occurs due to silica modification influence to the membrane’s pore which hinders these gas components (methane and nitrogen) to pass through the pores but allow other components (carbon dioxide and hydrogen) to diffuse. Carbon dioxide permeated faster through the pores of the membrane with the influence of surface diffusion mechanism. The obtained results are in good agreement with the literature [7]. Their findings were obtained by modifying macroporous alumina support with silica at pressures between 0.1 up to 6.0 bars and room temperature.
Figure 6: Gas flow rate across silica membrane after fifth dip-coating against feed pressure at 298 K.
Fig. 7 shows single gas permeation through the dip-coated silica membrane as a function of molecular weight at 1.0 barg and room temperature. The membrane exhibits a linear dependence on the inverse square root of molecular weight which corroborates with the literature [25] as expected for a process dominated by Knudsen diffusion mechanism.
Figure 7: Single gas permeation against molecular weight across silica membrane at 1 barg and room temperature.
Fig. 8 shows the selectivities of CO2 over N2, H2, O2, CH4 and He against feed pressure across
silica membrane at room temperature. CO2 selectivity was obtained using the ratio of permeability
of CO2 to that of each of the other gases (N2, H2, CH4, O2 and He) at the same pressures and
temperatures using equation (5). The ideal Knudsen selectivity value was obtained from the inverse square root of the ratio of the respective gas molecular weight. It can be seen that CO2 selectivity
decreased with pressure increase. However, CO2/CH4 selectivity of 24.07 was obtained at 0.7 barg
and 298 K. Also, the selectivity values obtained for this experiment were higher than the ideal
Knudsen separation.
Figure 8: CO2 selectivity against feed pressure across silica membrane at room temperature.
M. Asaeda & S. Yamasaki (2001) [25] applied the sol-gel techniques to fabricate thin layer silica membranes on porous silica and silica-zirconia supports coated on α-alumina porous cylindrical tubes. The pore size of the silica membrane was around 0.35 nm. They claimed that CO2 permeance
increased through the silica membrane as the temperature decreased, while N2 and CH4 permeances
increased very slightly, as CO2 is more adsorptive on the silica surface than N2 or CH4. They
reported that the porous silica membranes were quite stable when used in dry conditions, while a silica membrane on a silica-zirconia sub-layer was even stable in humid conditions. They achieved a selectivity of 25 for CO2/CH4 at 300 0C which is comparable to this experiment.
Conclusion
The dip-coated membrane prepared in this experiment exhibit high CO2 selectivity for pressures
up to 5.0 barg from other gas components. The higher CO2 permeance rate is attributed to the
deposition of the silica layer which resulted in the surface diffusion mechanism. Single gas experiment exhibits a linear flow relationship with respect to the inverse square root of molecular weight at room temperature and 1.0 barg as expected for a process dominated by Knudsen diffusion mechanism. CO2 selectivity value of 24.07 was also obtained for CO2/CH4 at 0.7 barg and room
temperature. Such a selectivity value could be useful in small-scale carbon dioxide removal unit for natural gas treatment processes.
Acknowledgement
The authors gratefully acknowledge Petroleum Technology Development Fund (PTDF) Nigeria for funding this research; Centre for Process Integration and Membrane Technology, (CPIMT), School of Engineering, The Robert Gordon University, Aberdeen, United Kingdom for providing the ceramic membrane and the reactor, and School of Pharmacy & Life Sciences RGU Aberdeen for the SEM results.
References
[1] Y. Yildirim, & R, Hughes. An experimental study of CO2 separation using a silica based
composite membrane. Trans IChem, 81, B, pp. 257-261, 2003.
[2] http://www.telegraph.co.uk/earth/environment/climatechange/9717280/Doha–Latest-figuressh-ow-global-CO2emissions- are-rising.html [Accessed 10th February 2013].
[3] International Energy Outlook 2013, http://www.iea.org/Textbase/npsum/WEO 2013 SUM.pdf. [Accessed 10th February 2013].
[4] M. R. Othman, H. Mukhtar, & A. L. Ahmad. Gas permeation characteristics across nano-Porous inorganic membranes. IIUM Engineering Journal, vol. 5, pp. 17-33, 2004.
[5] M. N. Kajama, N. C. Nwogu, & E. Gobina, Gas Permeation Study Using Porous Ceramic Membranes. Energy and Environment Research journal, vol. 4(3), pp. 43-49, 2014.
[6] M. Mulder. Basic Principles of Membrane Technology, Second Edition, Centre of Membrane Science and Technology. Kluwer Academic Publishers, University of Twente, Enschede, The Netherlands, pp. 227, 1996.
[7] N. C. Nwogu, E. Gobina, & M. N. Kajama. Improved carbon dioxide capture using nanostructured ceramic membranes. Low Carbon Economy, vol. 4, pp. 125-128, 2013.
[8] Y. Yildirim, & R. Hughes. The Efficient Combustion of O-xylene in a knudsen Controlled Catalytic Membrane Reactor. Trans IChemE. 80(Part B): pp. 159-164, 2002.
[9] L. Zhang, I. Park, K. Shqau, W.S. Winston Ho, & H. Verweij. Supported inorganic membranes. Promises & Challenges, Vol. 61 pp. 61-71, 2009.
[10] A. L. Ahmad, & N. N. N. Mustafa. Sol-gel synthesized of nanocomposite palladium-alumina ceramic membrane for H2 permeability: preparation and characterization. International Journal
of Hydrogen Energy, Vol. 32 pp. 2010-2021, 2007.
[11] http://ufdcimages.uflib.ufl.edu/AA/00/00/03/83/00141/AA00000383_00141_00062.pdf. [Acc-essed on 22/05/2016].
[12] Kim, Y., Kusakabe, K., Morooka, S., & Yang, S. Preparation of microporous silica membranes for gas separation. Korean Journal of Chemical Engineering, 18, pp. 106-112, 2001.
[13] Ohwoka, A., Ogbuke, I., & Gobina, E. Performance of pure and mixed gas transport in reconfigured hybrid inorganic membranes part 2. Membrane Technology, pp. 7-9, 2012.
[14] C. A. Scholes, S. E. Kentish, & G. W. Stevens. Carbon dioxide separation through polymeric membrane systems for flue gas applications. Recent Patents on Chemical Engineering, vol. 1, pp. 52-66, 2008.
[15] Wall, Y., Ompe-Aime-Mudimu., Braun, G., & Brunner, G. Gas transport through ceramic membranes under super-critical conditions. Desalination, 250, pp. 1056-1059, 2010.
[16] Jae-Hyun So, S.M. Yang & S. B. Park. Preparation of silica–alumina composite membranes for hydrogen separation by multi-step pore modifications. Journal of Membrane Science, Vol. 147, pp. 147-158, 1998.
[17] F. I. Khan, & A. K. Ghoshal. Removal of volatile organic compounds from polluted air. Journal of Loss Prevention in the Process Industries, 13(6), pp. 527-545, 2000.
[18] A. F. M. Leenaars & A. J. Burggraaf. The Preparation and Characterization of Alumina Membranes with Ultrafine Pores: Part 3. The permeability of pure liquids. Journal of Membrane Science, Vol. 24, pp. 245-260, 1985a.
[19] Smart, S. Liu, S. Serra, J. M. Diniz da Costa, J. C. Iulianelli, A. & Basile A. Porous ceramic membranes for membrane reactors. 2013.
[20] G. Q. Lu, J. C. Diniz da Costa, M. Duke, S. Giessler, R. Socolow, R. H. Williams, & T. Kreutz. Inorganic membranes for hydrogen production and purification: A critical review and perspective. Journal of Colloid and Interface Science, vol. 314, pp. 589-603, 2007.
[21] J. D. Way, & D. L. Roberts. Hollow Fiber Inorganic Membranes for Gas Separations. Sep. Sci. Tech., 27(1), pp. 29-41, 1992.
[22] E. Gobina, Apparatus and Methods for Separating Gases. United States Granted Patent No. US 7048778, May 23, 2006.
[23] P. Huang, N. Xu, J. Shi, & Y. S. Lin. Recovery of volatile organic compounds from air by ceramic membranes. Ind. Eng. Chem. Res., vol. 36, pp. 3815-3820, 1997.
[24] A. Ohwoka, I. Ogbuke, & E. Gobina. Performance of pure and mixed gas transport in reconfigured hybrid inorganic membranes part 1. Membrane Technology, pp. 7-12, 2012.
[25] B. A. McCool, N. Hill, J. DiCarlo, & W. J. DeSisto. Synthesis and characterization of mesoporous silica membranes via dip-coating and hydrothermal deposition techniques. Journal of Membrane Science, vol. 218, pp. 55-67, 2003.
[26] M. Asaeda, & S. Yamasaki. Separation of Inorganic/Organic Gas Mixtures by Porous Silica Membranes. Sep. Purif. Tech., 25, pp. 151-159, 2001.
View publication stats View publication stats
