Poly ether block Amide/Polyvinyl Alcohol/ MgO nanocomposite membranes: Selectivity for CO₂ and CH₄ gases

2.1 Materials

The polymeric constituents utilized were polyether block amide branded Pebax-2533(PEBA, Pebax 2533) and polyvinyl alcohol (PVA), purchased from Arkema (France) and Merck (Germany) respectively. Magnesium oxide (MgO) nanoparticles (particle size < 50 nm) and 95 % purity were obtained from USNANO to formulate the nanocomposite membrane and acetic acid (glacial, Reagent Plus ®, ≥99 %) were purchased from Sigma Aldrich. As solvent, dimethylacetamide was selected - a colorless polar liquid. Chemical properties of the materials are delineated in Table 1.

2.2 Methods

2.3 Membrane characterization

2.3.4

2.3.4 Molecular dynamics simulation

Molecular dynamics is a method to observe atomic and molecular motions on the microscopic scale. According to Newton's formula, MD allows the examination of microscopic movements of systems containing N atoms(Bahreini et al., 2024, 2023):

(6)

$$m_i\frac{{\partial }^2{r}_i}{\partial t^2}=F_i,i=1\cdots N.$$ Eq. (7) states that the forces (Fi) are equivalent to the potential function (V), in which ri and t stand for the center-of-mass position vector related to penetrant i and the time, respectively.

(7)

$$F_i=-\frac{\partial V}{{\partial r}_i}$$ The equations are then solved in small time steps at the same time. The system is monitored for some time while regulating the temperature and pressure, and the coordinates are sent to an external file at scheduled intervals. A represented trajectory of the system is these coordinates as a function of time. Initially, the system typically achieves equilibrium. Numerous microscale features, like diffusion processes in membranes, may then be retrieved from the output file by calculating the average over an equilibrium trajectory. The diffusion coefficient of a gas in an organic solvent, polymer, or clay may be computed by running an MD simulation and calculating the gas's mean square displacement (MSD) in the media. Assuming that the penetrant molecules exhibit random walks in the polymer matrix over extended periods, the mean-square-displacement factor is linear with time, and the diffusion coefficients may be determined utilizing the Einstein formula(Ullal and Spearot, 2014):

(8)

$$D=\frac{1}{6N}\underset{t\to \infty }{\lim }\frac{d}{\mathit{dt}}\langle \sum _i^N{\left[r_i\left(t\right)-r_i\left(0\right)\right]}^2\rangle$$ Tracking the penetrant's migration |r(t)-r(0)| from its initial location allows us to study the motion profile of penetrant gases in the substrate polymer. After that, the slope of a plot of the mean squared displacement versus time t is used to determine the diffusion coefficient D.

(9)

$$\langle \sum _i^N{\left[r_i\left(t\right)-r_i\left(0\right)\right]}^2\rangle$$ In this regard, the PEBA and PVA were modeled by putting together ten chains of 30 monomer units as shown in Fig. 1.

Close

Fig. 1

Pebax® 2533 and PVA structures.

Export to PPT

The polymer model for Pebax® 2533 and PVA contains 5370 and 2120 atoms, respectively, and the sides of the cubic unit cells measure approximately 29–31 Å. Using the Materials Studio suite of software's Amorphous Cell module, several configurations of polymer chains (depicted in Figure B of the Appendix) were produced and arranged in a cubic space based on Theodorou and Suter's “self-avoiding” random-walk method and the Meirovitch scanning technique, employing periodic boundary conditions (PBC) to model the cubic structure(Meirovitch, 1983). At 1 bar and 298 K, the aim was to generate refined systems approximating the experimental density of the amorphous polymers. In all simulations, Berendsen's method [40] controlled temperature and pressure using a half-life of 0.1 ps (decay constant) and a pressure scaling constant of 0.1 ps, with a system compressibility of 0.5 GPa-1.

Four CO₂ and CH₄ solute molecules were placed at the center of the simulation space comprising 10 chains of 20 monomers each of the PEBA and PVA blend, to model gas diffusion. A group-based cutoff was employed in this investigation. The COMPASS forcefield described all interactions, using a cutoff value of 12.5 Å. A 3 Å width spline function interpolates the nonbonding interactions from their full value to zero within this range due to computational constraints. Both the cutoff and splining were also modified for long-range effects. For Geometry Optimization the Maximum Iterations was set to 1000.

3.1 Fourier-transform infrared spectroscopy (FT-IR)

FT-IR was employed to characterization and identify the functional groups in the membranes of composite and nanocomposite structures. The results of this investigation for the specimen membrane are presented in Fig. 2, respectively. In Fig. 2, the observed peaks for the Pebax in the range of 3300 and 2860 cm⁻¹ correspond to the stretching vibrations of N–H and aliphatic C–H groups, respectively. Additionally, absorption peaks appearing in the regions of 1720, 1630, and 1100 cm⁻¹ correspond to the C = O, H-N-C = O, and C-O bonds, respectively. The curves in Fig. 2 reveal peaks for PVA in the range of 3300–3500 cm⁻¹, indicative of the stretching vibrations of O–H bonds resulting from intermolecular and intramolecular hydrogen bonding. Moreover, observed absorption peaks in the regions of 1742 and 1140 cm⁻¹ are associated with the C = O and C-O bonds arising from the remaining acetate groups in the structure. The absorption peak at 1627 cm⁻¹ indicates carbonyl groups, while the peak at 1147 cm⁻¹ signifies the presence of crystalline regions in the PVA chain structure formed due to symmetric stretching of C–C or C-O bonds resulting from intramolecular hydrogen bonding. As seen in Fig. 2, in the FTIR spectrum of the Pebax/PVA membrane, the positions of the peaks have slightly changed due to the interaction between the Pebax and PVA. This can be attributed to the suitable interaction between the carbon and nitrogen atoms of the Pebax and the hydrogen atoms of PVA. The positional shift of H-N-C = O in the Pebax from 1600 to 1630 cm⁻¹ in the Pebax/PVA composite membrane is observed. Furthermore, the stretching vibration of O–H in the polyether block amide has shifted from 3300 to 1325 cm⁻¹ in the mixed Pebax/PVA membrane(Klepić et al., 2020; Martínez-Izquierdo et al., 2019; Mina Kheirtalab, Reza Abedini, 2022; Wong et al., 2021).

Close

Fig. 2

FTIR spectra of membranes.

Export to PPT

3.2 X-ray diffraction

The interaction between nanoparticles and polymer chains may cause the nanoparticles to change crystalline form. The change in crystal size and form can be detected by an X-ray diffraction test. The structures of Pebax 1657 polymer, Pebax/PVA composite membrane, and Pebax/PVA/MgO nanocomposite membrane were investigated using X-ray diffraction (XRD). Fig. 3 illustrates the X-ray diffraction patterns of the fabricated membranes. All polymers exhibit prominent peaks in the range of around 24°, indicative of small, distributed crystals within the polymeric membrane composition. With the addition of 10 % by weight of MgO nanoparticles to the polymeric blend of Pebax/PVA, a reduction in peak intensity in the 24° regions is observed. The decreasing trend in the composite and nanocomposite membranes suggests a reduction in the crystallinity of the membranes(Kheirtalab and Reza Abedini, 2022; Kheirtalab and Reza Abedini, 2020). Moreover, the addition of MgO nanoparticles to the polymer blend has led to the emergence of peaks in the 42° and 62° regions. This phenomenon could be attributed to the placement of MgO nanoparticles between the polymer chains, preventing them from closely packing and linking polymer particles together. On the other hand, the addition of MgO nanoparticles has resulted in broader peaks, indicating a decrease in crystalline properties and facilitating the flexibility of the nanocomposite membrane. In fact, MgO, by reducing the crystalline portion of the membranes and enhancing the polymer chain flexibility, contributes to the improvement of the gas transport properties of CO₂. Interaction between nanoparticles and polymer chains may cause changes in the crystalline shape and size of the nanoparticles, which can be detected through X-ray diffraction experiments.

Close

Fig. 3

XRD graph of optimal formulations.

Export to PPT

3.3 Study of morphology

The morphological investigation of composite membranes with different ratios, including Pebax100/PVA0%, Pebax80/PVA20%, Pebax50/PVA50%, and Pebax20/PVA80%, was conducted using field-emission scanning electron microscopy (FE-SEM). FE-SEM images of the fabricated membranes are presented in Fig. 4. a-c. As observed in Fig. 4. a, Pebax membranes exhibit a texture structure with irregular distribution, and spherical-shaped pores are also visible(Nobakht and Abedini, 2022). In the images of Pebax80/PVA20% composite membrane (Fig. 4. b), a heterogeneous structure with irregular distribution is evident. This heterogeneity can be attributed to the immiscibility of Pebax and PVA membranes with each other. Furthermore, in Fig. 4.c & d, as the PVA content increases, the cross-sectional surface of the samples becomes smoother due to improved compatibility between the two polymers. With an increase in PVA content, Pebax chains are adequately dispersed, forming a homogeneous phase. The more uniform phases lead to increased gas permeability and enhanced selectivity due to reduced entanglement. The smooth surface and improved compatibility between the polymers contribute to the dispersion of Pebax chains, forming a more uniform phase(Chen et al., 2023). This uniformity results in enhanced gas permeability and increased selectivity due to reduced entanglement.

Close

Fig. 4

FE-SEM micrographs of samples containing a) 0, b) 20, c) 50 and d) 80 % of PVA.

Export to PPT

In the images of Pebax20/PVA80/MgO15% and Pebax80/PVA20/MgO10% nanocomposites in the Fig. 5, it is evident that the addition of MgO nanoparticles during the membrane preparation stages enhances the texture and structure of the membrane compared to the Pebax/PVA composites. The surface roughness increases, and the nanoparticles are well-dispersed in the polymer matrix. The resulting membrane exhibits a uniform distribution with a coherent structure and regular holes, contributing to the overall improvement in membrane performance. The images also reveal that the nanoparticle size for Pebax20/PVA80/MgO15% is 302 nm, and for Pebax80/PVA20/MgO10% is 285 nm. This increase suggests swelling, exfoliation, and dispersion of nanoparticles within the polymer chains. Additionally, the strong interaction between nanoparticles and the polymer substrate, coupled with the reduction of the crystalline phase, is observed as the driving force behind the increased permeability and gas separation efficiency.

Close

Fig. 5

Nanoparticles’ dispersion in polymer matrix for a) Pebax20/PVA80/MgO15%, and b) Pebax80/PVA20/MgO10%.

Export to PPT

3.4 Analysis of gas

According to previous studies, CO₂ permeability increases with increasing pressure in polymer membranes. While for light gases such as CH₄, this trend is decreasing or unchanged. According to the results of the investigations, the permeability of CO₂ gas is higher than that of CH₄. The greater permeability of CO₂ compared to CH₄ in Pebax/PVA membrane can be attributed to the interaction of CO₂ and the softening effect of CH4 on the polymer. In addition, another reason for the higher permeability of CO₂ compared to CH₄ is the high critical temperature of CO₂ compared to CH₄ gas. In terms of molecular size, carbon dioxide is smaller than the studied methane gas and has a higher condensability. The results of CO₂ and CH₄ gas permeation at pressures of 2, 4, 6 and 8 bar and temperatures of 25, 35, 45 and 55 °C by Pebax/PVA mixed membrane with different weight ratios of PVA are shown in Figs. 6 and 7. As can be seen in Figs. 5 and 6, gas permeation values through the membrane increase with increasing pressure due to increasing driving force caused by mass transfer and solubility of gases. By increasing the pressure from 2 to 8 bar, the permeability of CO₂ and CH₄ gas has increased. In the rubber membrane, the permeability of light and non-polar gases such as methane is independent of pressure and remains constant with increasing pressure, but the permeability of condensable gases increases with increasing pressure. This increase can be related to the greater affinity of CO₂ with the polar ether groups of Pebax, which increases the CO₂ absorption in the membrane with the increase in pressure, followed by an increase in solubility. The investigation of changes in permeability with temperature for the Pebax/PVA nanocomposite membrane is shown in Fig. 6-b and Fig. 7-b. As can be seen, the permeability increases as the temperature increases to 55 degrees Celsius. The temperature dependence of permeation for CO₂ and CH₄ gases in the membrane is probably due to the double effect of permeation and solubility(Feng et al., 2023).

Close

Fig.6

A) the effect of pressure on the permeability of co₂ gas and b) The effect of temperature on the permeability of CO₂ gas of samples containing 0, 20, 50, and 80 % of PVA.

Export to PPT

Close

Fig.7

A) the effect of pressure on the permeability of ch₄ gas and b) The effect of temperature on the permeability of CH₄ gas of samples containing 0, 20, 50, and 80 % of PVA.

Export to PPT

The measured CO₂ permeability of the membranes was between, 53.569 to 315.440. The most suitable model to fit the relationships between the individual variables and Carbon dioxide permeability is a derived quadratic, according to the analysis of variance (ANOVA) (Table C in Appendix). The model's F-value (19.26) suggests that it is significant. Figure C in the Appendix compares the experimental results to predicted CO₂ permeability values. The normalized graphs of the findings are provided in graph a in Figure C in the Appendix for faster and more effective comparisons. Based on the actual value of the findings, graph b in Figure C in the Appendix presents the model's anticipated value of the test results. The residual value in terms of formulations is shown in graph c in Figure C in the Appendix (16 formulations in this study). The residual value (the deviation for both the actual values and the values obtained from the model) is represented in terms of the predicted value in graph d in Figure C in Appendix. The Percentage error for Carbon dioxide permeability among all experimental tests varied between −5.33 percent and +5.33 percent, as seen on the graph's y-axis. As a result, the model is best suited because the discrepancy across the experimental and anticipated values is quite small. The formulation should be skipped if one of the effects falls beyond this range.

Fig. 8 depicts the results of the CO2 permeability test. As illustrated in Fig. 8, the effect of PVA percentage on CO₂ permeation is slightly complex and differs from high pressure and temperature at low pressures and temperatures. Two factors control the penetration of gas into the membrane. One is the polar interactions between CO₂ and PVA molecules that cause more gas molecules to dissolve inside the membrane, thereby increasing the gas permeation. On the other hand, with increasing the percentage of PVA, the possible crystal points inside the membrane increase, which as a barrier prevents the penetration of gas from the membrane. The outcome of these two factors determines the gas permeation through the membrane according to the conditions. MgO nanoparticles also have a great effect on decreasing the CO₂ permeability (graph a in Fig. 8), MgO nanoparticles are on the nanometric scale, interact with polymer chains greatly and act as a barrier18. Fig. 8 illustrates that the CO₂ permeability increases as the gas pressure increases. This event indicates that tiny molecules of gases were capable of entering and transferring via the polymer chain, increasing CO₂ permeability. By increasing the gas temperature, molecular movements increase and cause more gas to transfer through the membrane. In addition, the warmer gas raises the membrane temperature and increases the free volume of the polymer chains. Gas molecules pass more easily through polymer chains with higher free volume.

Close

Fig. 8

A response surface plot depicting the influence of A to D variables on CO₂ permeability.

Export to PPT

As the study shows that the independent variables influence CO₂ permeability. Eq. (10) displays the mathematical correlation for both CO₂ permeability and the A to D variables acquired from the regression study.

The results of examining the selectivity of CO₂ and CH₄ gas at pressures of 2, 4, 6, and 8 bar and temperatures of 25, 35, 45, and 55 °C by Pebax/PVA mixed membrane with different weight ratios of PVA are presented in Fig. 9. As can be seen, permeability and selectivity increase at higher pressures due to the strengthening of the driving force and softening effect. Examining the values obtained from the Pebax/PVA nanocomposite membrane in the selectivity study showed that, as a result of the compatibility of the polymer surface, the polymer chains can create gaps or narrow seams due to the strong interaction around PVA, which increases the selectivity. Another reason for the simultaneous increase in permeability and selectivity is the absence of non-selective holes in the membrane structure. An important point to note is that due to the selectivity of glass polymers, the presence of polyvinyl alcohol polymer in the Pebax membrane increases the selectivity(Mina Kheirtalab, Reza Abedini, 2022, 2020).

Close

Fig. 9

The effect of nanoparticles on the permeability of CO₂ gas in Pebax/PVA/MgO nanocomposite membrane of samples containing 0, 20, 50, and 80 % of PVA.

Export to PPT

The measured CH₄ permeability of the membranes ranged from 12.77 to 35.22 (units). According to the analysis of variance (ANOVA) (Table C in the Appendix), the most suitable model to explain the relationship between the independent variables and CH₄ permeability is a modified one. The model's F-value (8.36) indicates its statistical significance. Figure D in the Appendix compares the experimental findings with the predicted values for CH4 permeability. For more detailed and precise comparisons, the normalized graph of the findings is provided in graph a of Figure D in the Appendix. Graph b in Figure D in the Appendix depicts the model's predicted values based on the actual findings. The residual values in terms of formulations are illustrated in Graph c of Figure D in the Appendix, while the residual values (deviation between the observed and predicted values) are shown in terms of the predicted values in graph d of Figure D in the Appendix. The relative error for CH4 permeability of all experimental runs fluctuated within −5.33 % and + 5.33 %, as indicated on the y-axis of the graph(Sambasevam et al., 2023).

Fig. 10 illustrates the results of the CH₄ permeability test. In contrast to the effect of PVA on CO₂ permeability, increasing PVA content reduces CH₄ permeability. This can be attributed to the reduced interactions between CH₄ and PVA molecules, resulting in an increased barrier effect of PVA crystals against CH₄ molecules. The effects of MgO nanoparticles, pressure, and temperature on CH₄ permeability are similar to their effects on CO₂ permeability(Zou et al., 2023).

Close

Fig. 10

A response surface graph depicting the influence of A to D variables on CH₄ permeability.

Export to PPT

As the study shows that the independent variables influence CH₄ permeability. Eq. (11) exhibits the quantitative correlation between CH₄ permeability and the A to D variables derived from the regression study.

3.5 Molecular dynamics simulation

Comparison between experimental results and simulation data validates the reliability of experimental findings. Fig. 11 plots mean squared displacement (MSD) versus time, allowing diffusion coefficients to₂ be determined per Equation (5). This quantifies displacement of test gases from original lattice positions [ri(t) - ri(0)] during dynamic motion within the host polymer. Dividing the MSD curve slopes by 6*10⁴ calculates individual gas emission coefficients(Sateria et al., 2023). Table 3 reports diffusion coefficients for CH₄ and CO₂ through membranes featuring varied PVA loadings. The data indicates CO₂ diffusion coefficients rise with increasing PVA content, whereas CH₄ values decrease. This performance trend links to membrane polarization at higher PVA proportions. As CO₂ molecule polarity matches the enriched surface environment more strongly, it preferentially permeates the material. Since selectivity relates to the diffusion coefficient ratio between gases, incorporating more PVA lowers selectivity for CH₄/CO₂ separation as a consequence of the described phenomenon.

Close

Fig. 11

MSD (Å) vs Time (ns) for a) CH₄ and b) CO₂ at different percentages of PVA.

Export to PPT

Fig. 12 shows the evaluation of CO₂ and CH₄ diffusion coefficients obtained from molecular simulation data and gas permeability from experimental measurements at varied PVA weight percentages in dispersed polymer nanocomposites. In Fig. 12. a, the CH₄ emission coefficient from simulation decreases non-linearly with a gentle slope as PVA content rises. Fig. 12. c demonstrates experimental CH₄ permeability mirrors this trend. This occurs as higher PVA proportions enhance the material's polarity, hindering dissolution of nonpolar CH₄. Meanwhile, Fig. 12. b illustrates the CO₂ diffusion coefficient from simulation data increases slightly with growing PVA percentage, resembling the experimental CO₂ permeability change in Fig. 12 at low pressures and temperatures. Carbon dioxide's polar vibrational nature facilitates greater penetration into the polymer matrix where PVA loading is elevated.

Close

Fig. 12

Gas diffusion coefficients for a) CH₄ and b) CO₂ obtained from molecular dynamics simulation and c) CH₄ and d) CO₂ gas permeabilities obtained from experimental data.

Export to PPT

3.6 Optimal formulations

Further variations in gas permeability compared to each other can serve as a useful criterion for assessing the selectivity of nanocomposite membrane permeation in relation to CO₂ or CH₄ gases. Equation (12) presents the relative separation of CO₂ and CH₄ gases by the nanocomposite membrane, where P_CO2 represents the CO₂ permeability and P_CH4 represents the CH₄ permeability(Jasim et al., 2023).