Gas permeation,thermal,morphology and mechanical properties of polyimide/clay nanocomposites: Effect of organically modified montmorillonite

Abstract

Novel polyimide/clay nanocomposites were fabricated by incorporating organically modified clay at different loading percentage of 1–5 wt% within the polyimide (PI) matrix. A novel PI was synthesized through a direct polycondensation reaction between novel 5,5′-((4-methoxyphenyl)azanediyl)bis(isobenzofuran-1,3-dione)and 4,4′-(1,4-phenylenebis(oxy))bis(3-(trifluoromethyl)aniline). The organoclay was prepared from Sodium montmorillonite (Na-clay) and protonated form of diamine (modifier) via ion-exchange reaction. The structural characterization of nanocomposites were made using Fourier transform infrared spectroscopy (FT-IR), X-ray diffraction (XRD), scanning electron microscopy (SEM) and transmission electron microscopy (TEM), the mechanical properties were obtained by tension tests. The results of XRD, TEM, and FE-SEM indicated that clay could well disperse in the PI matrix and were intercalated by diamine and PI macromolecules. The thermal stability and glass transition temperature of nanocomposite at lower clay loadings are similar to PI but gradually decrease with increase in clay content. The tensile properties of PI/Clay nanocomposites showed the highest values at the optimum clay content of 3 wt%. The gas permeation of PI/clay nanocomposites was increased with the increase in loading percentage of organoclay, whereas the gas selectivity was maintained at par to that of neat PI film until it diminished at 5 wt% percentage of organoclay causing mineral agglomeration. PI/clay 3 wt% MMMs exhibited 245% and 240% increase in CO₂ and O₂ gas permeability respectively, while showing minimal variation in the selectivity of O₂/N₂ and CO₂/N₂ compared to that of neat PI matrix.

Keywords

Polyimide/clay nanocomposites polycondensation gas permeation gas selectivities

Introduction

The unrelenting efforts to form novel materials to surpass the ‘upper bound’ limitation lead to the development of state-of-theart organic–inorganic hybrids, generically called mixed matrix membranes (MMMs). MMM is comprised of discrete inorganic phase dispersed in a continuous polymer phase which brings forth novel hybrid material with synergistic advantages from each phase: high selectivity of the dispersed particle, desirable mechanical integrity, and economical processability of the polymer matrix.^1–3 At the early stage of MMM development, selective adsorbent or molecular sieve inorganic particles such as zeolite became the main choice for the filler material.⁴

Polyimides (PIs) are widely used in the microelectronics industry as buffer coatings, passivation layers, α-particle barriers, interlayer dielectrics, wafer-scale packages, etc., because of their mechanical, thermal and electrical properties, and chemical resistance. In order to improve the utility of PIs in electronic industries, many kinds of PI/clay nanocomposites have been studied using different dianhydrides and diamines. Various dianhydrides widely used to prepare PI/clay nanocomposites are pyromellitic dianhydride (PMDA), 3,3′,4,4′-biphenyltetracarboxylic dianhydride (BPDA), benzophenone-3,3′,4,4′- tetracarboxylic dianhydride (BTDA), 2,2-bis [4-(3,4-dicarboxyphenoxy)phenyl]propane dianhydride, oxydiphthalic anhydride (ODPA) and 2,2-bis(3,4-dicarboxyphenyl) hexafluoropropane dianhydride (6FDA).^5–10

Clay minerals have been shown to be good candidates for the preparation of nanocomposites because of their easy availability, low cost, layered structures and ability to become organophilic by cation-exchange reaction. Clay minerals such as montmorillonite, saponite and hectorite have mainly been used to form polymer-based nanocomposites.¹¹ These clayminerals usually contain hydrated inorganic cations, such as Na⁺, K⁺, Ca²⁺, and are hydrophilic. These clay minerals have to be rendered more organophilic to make them compatible with organic polymers. To attain this, the hydrated cations of clay minerals are exchanged with cationic surfactants, such as quaternary alkylammonium cations.¹²

On the one hand, the molecular diffusion and shear critically depend on the surface modification of montmorillonite (MMT), such as an optimal structure, size and adsorption content of surfactant. The organic modification of MMT decreases the energy of adhesion between layers, thus facilitating the molecular diffusion and shear.^13,14 The increase of the number of alkyl tails or the adsorption content of surfactants will increase the interlayer distance and weaken the attraction between adjacent platelets.^4,15 However, it also increases steric hindrance, and the molecular diffusion of polymers becomes more difficult.

The chemical structure and melt viscosity of polymers influence the molecular diffusion and shear. Balazs proposed a structure of polymer that was beneficial to the exfoliation via molecular diffusion.^13,16 The polymer should contain fragments attracting to the MMT surface in order to promote the compatibility between the polymer and MMT; meanwhile the polymer should also contain longer fragments that are not attracted to MMT surface in order to exfoliate MMT layers. As we all known, polar polymers have stronger interaction with the surface of MMT than polar polymers.^17–19 The favorable interaction makes polymers intercalate into the galleries of OMMTs easily, and shear stress can also be transferred effectively. But the resistance to diffusion of polar polymers in the interlayers also increases with the enhancement of interactions. Therefore the exfoliation cannot be achieved via molecular diffusion in some polar matrixes.²⁰ In addition, the melt viscosity of polymer is also a vital factor because different melt viscosities will result in different diffusion ability and shear stress. The exfoliation needs a balance between shear stress that requires high viscosity and diffusion process that needs rather low viscosity.^21,22

In order to synthesize novel high performance materials with improved processability and solubility with respect to aromatic PI, PI nanocomposites were prepared through a solution intercalation method using different loading percentage organoclays. The effects of nanofiller type on morphology, thermal properties, gas permeation and mechanical of PI were explored in details.

Materials

5-bromo-2-methylisoindoline-1, 3 dione, 4-methoxyaniline and tetrakis(triphenylphosphine) palladium (0) were purchased from Alfa Aesar and used as received. Commercially available N-methyl-2-pyrrolidinone (NMP), m-cresol, N, N-dimethylacetamide (DMAc), N,N-dimethylformamide (DMF), dimethyl sulfoxide (DMSO), tetrahydrofuran (THF) and other reagents were all used as received. Research grade ethane, ethylene, propane and propylene gases (purity499.95%) were purchased from Technical Gas Service. All the materials and solvents that were used were obtained from the suppliers without further purification. The sodium montmorillonite clay (Closite Na+, cation-exchange capacity (CEC) = 92 mmol/100 g, denoted as Na-MMT) was purchased from Southern Clay Products Co. (USA).

Equipments

Fourier transform infrared spectroscopy (FT-IR) analysis was performed on an FT-IR spectrometer (ALPHA, Bruker) in the frequency range of 400–4000 cm-1 at room temperature. Xray diffraction (XRD) patterns were recorded on a diffractometer (D/max-2500, Rigaku), using CuKα radiation at a scanning rate of 5°/min (from 10° to 80°). Proton nuclear magnetic resonance (¹H NMR, 500 MHz) spectra were carried out using Bruker Avance 500 instrument (D-755 of the Roger-Gaudry, Germany) at room temperature (RT) in dimethylsulphoxide-d6 (DMSO-d₆). Multiplicities of proton resonance were designated as singlet (s), doublet (d), triplet (t), and multiplet (m). The microstructure of PI/mMMT nanocomposites were measured by scanning electron microscope (FE-SEM) (EVO18, Zeiss). In order to maintain the original appearance of PI/m-MMT nanocomposites, the samples were treated by liquid nitrogen to obtain their fracture surface. The spatial dispersion of m-MMT in polymer matrix was analyzed via TEM, transmission electron microscopy, FEI Tecnai F20S. Thermal gravimetric analysis (TGA) was performed on a Setaram TGA 92 Thermobalance, by heating the clay samples with 10°C/min heating rate from ambient temperature to 800°C at nitrogen atmosphere. The gas permeation performance of PI I/m-MMT nanocomposite membranes were carried out by using 99.999% pure CO₂ O₂and N₂ gases at room temperature at a pressure of 2–10 bars. The desired gas was supplied through lumen side of the membranes bundle. For permeation test, 3 fibres are taken as a pile and one end of the pile was sealed using epoxy glue. After 3 h, another coat of epoxy was performed at pile, followed by 24 h of drying time. Subsequently, after 24 h, pile of fibres are fixed inside the module and glued again so that no gas can pass through except through the fibres shell side. The permeation cell was vacuumed prior to gas permeation test. Membrane permeance and selectivity for CO₂, O₂ and N₂ was calculated by using equations (1) and (2), respectively.^23,24

\frac{p_{i}}{l} = \frac{Q}{A Δ P} \frac{273.15}{T}

(1)

α_{i j} = \frac{\frac{p_{i}}{l}}{\frac{p_{j}}{l}}

(2)

Pi/l is the gas permeance in GPU [1GPU = 1 × 10⁻⁶ cm³ (STP)/cm².s.cmHg]. Subscript, i, represents the penetrating gas, i.e., CO₂ or O₂ and N₂. A, is the effective surface area of the membrane, ΔP is the pressure difference across membrane in cmHg and T (K) is the temperature at which permeation was carried out. α_ij, represents ideal selectivity of gas i to j.

Experimental

Monomer synthesis

The diimide compound 2 obtained from double N-arylation reactions of 4-methoxyaniline with 5-bromo-2-methylisoindoline-1, 3-dione was carried out in the presence of palladium (II) acetate, rac-2,2′-bis(diphenylphosphino) −1,1′-binaphthyl (rac-BINAP), and cesium carbonate in toluene. The resulting diimide compound (1) was then hydrolyzed with aqueous potassium hydroxide, giving the corresponding tetracarboxylic acid, which in turn was converted to the new dianhydride monomer 2 by the chemical cyclodehydration with acetic anhydride (Scheme 1). Elemental analysis, FT-IR, NMR and ¹³C-NMR spectroscopic techniques were used to identify structures of the dianhydride monomer. The carbonyl groups of diimide compound exhibit two characteristic bands at around 1720 and 1645 cm⁻¹ could be attributed to imide C = O asymmetric and symmetric stretching, respectively.²⁵ After hydrolyzed and cyclodehydrated to dianhydride monomer, the carbonyl groups shifted to around 1827 and 1717 cm⁻¹. Figures 1 and 2 illustrates the ¹H-NMR and ¹³C-NMR spectra of the dianhydride monomer 2 and these spectra agree well with the proposed molecular structure. Thus, the results of all the spectroscopic and elemental analyses suggest the successful preparation of the target dianhydride monomer 2.

Scheme 1.

Synthesis rout of novel diimide monomer.

Figure 1.

¹H-NMR spectrum of diimide monomer.

Figure 2.

¹³C-NMR spectrum of diimide monomer.

FT-IR (compound 2) (KBr, cm⁻¹): 3155 (w), 3055 (w), 2947 (w), 2922 (w), 1822(s), 1720 (s), 1645 (s) 1422 (m), 1348 (m), 1170 (w), 815(m), 725 (w).

¹H-NMR (compound 2) (500 MHz, DMSO-d₆, ppm): 2.48 (s, 3H, C-H), 7.44–7.46 (d, 2H, Ar-H, J = 4.0), 7.53 (s, 2H, Ar-H), 7.73–7.75 (d, 2H, Ar-H, J = 4.0), 8.22–8.24 (d, 2H, Ar-H, J = 4.0), 8.33–8.36 (d, 2H, Ar-H, J = 4.0).

¹³C-NMR (compound 2) (125 MHz, DMSO-d₆), δ (ppm): 56.27 (O-CH₃), 120.91 (Ar), 128.23 (Ar), 129.06 (Ar), 131.19 (Ar), 132.47 (Ar), 135.90 (Ar), 141.42 (Ar), 142.28 (Ar), 155.43 (Ar), 161.81 (C = O).

Elemental analysis calculated for theoretically (compound 2): For C₂₃H₁₃NO₇ (415.36 g mol⁻¹): C, 66.51%; H, 3.15%; N, 3.37%. Found for Synthesized monomer practically: C, 66.61%; H, 3.17%; N, 3.32%.

Polymer synthesis and preparation PI with amine end-capped

Novel PI was prepared by the one/pot, high/temperature solution polymerization of new dianhydride monomer with aromatic diamine (4,4′-(1,4-phenylenebis(oxy))bis(3-(trifluoromethyl)aniline) in m-cresol at 200°C in the presence of isoquinoline as the catalyst (Scheme 2). Polymerization reaction proceeded homogenously and led to the formation of highly viscous polymer solutions that can be precipitated into tough fiber-like forms when slowly trickling into methanol. The resulting PI exhibited inherent viscosities of 1.43 dL/g in N-methyl-2-pyrrolidinone (NMP) and can afford transparent/self/standing films via solution casting, indicating this is high molecular weight polymer. Moreover, the number average molecular weight (M_n), weight average molar weight (M_w) and polydispersity index (PDI) of the synthesized polymer was further supported by GPC measurements. The structure of these PI was also confirmed with IR and NMR spectroscopic techniques, and the spectra agree well with the proposed molecular structures. The IR spectra of PI exhibit characteristic imide absorption bands at around 1772 (imide asymmetrical C = O), 1725 (imide symmetrical C = O), 1352 (C–N), and 750 cm⁻¹ (imide ring deformation).^25,26 A typical ¹H NMR spectrum of PI shown in Figure 3 reveals that all the peaks could be readily assigned to the hydrogen atoms of the recurring unit.

Scheme 2.

Synthesis rout of novel PI and PI with amine end-capped monomer.

Figure 3.

¹H-NMR spectrum of polymer (PI).

PI was also characterized by elemental analysis techniques, and the results are in good agreement with the calculated ones for the proposed structures. These results in sum confirmed the successful formation of the new PI.

FTIR (KBr, cm⁻¹): 3,455 (m, br), 3,150 (w), 2,905 (w), 1,772 (m), 1,725 (s), 1,655 (m), 1,550 (s), 1,477(s), 1,385 (s), 1,352 (s), 1,319 (s), 977 (s), 950 (m), 845 (m), 725 (m), 715 (m), 655 (m), 540 (m).

¹H-NMR (400 MHz, DMSO-d₆, ppm): 4.14 (s, 3H, C-H), 7.09–7.11 (d, 2H, CH, J = 3.6 Hz), 7.10–7.12 (d, 2H, Ar-H, J = 6.4 Hz), 7.14–7.16 (d, 2H, Ar-H,J = 8.4), 7.18–7.21 (d, 2H, Ar-H,J = 4.0), 7.22–7.23 (d, 2H, Ar-H,J = 4.0), 7.25 (s, 2H, Ar-H), 7.28 (s, 2H, Ar-H), 7.29 (d, 2H, Ar-H,J = 4.0), 7.30–7.32 (d, 2H, Ar-H, J = 6.2).

GPC measurements of resulting PI: M_n: 88 146, M_w: 199 089, PDI: 2.25

Elemental analysis calculated for C₄₃H₂₃F₆N₃O₇ (809.05 g mol⁻¹): C, 63.79%; H, 3.11%; N, 5.19%, F, 14.08. Found for Synthesized monomer practically: C, 63.05%; H, 3.17%; N, 5.32%, F, 14.21.

Fabrication of organoclay and PI/Clay nanocomposites

The organoclay was prepared by a cation-exchange method, which is a displacement of the sodium cations of Na-clay with the quaternary ammonium ions of the surfactant. Typically, 10 g of Na-clay was mechanically stirred with 500 mL of deionized water at 25°C for 3 h to swell the clays. Then, 15 mL protonized diamine/HCl solution was added dropwise into the Na-MMT suspension. The mixed suspension was stirred at 65°C overnight to obtain the organoclay layers. The precipitates of organoclays were isolated by suction-filtration using Whatman filter paper placed in a 500 mL beaker again and washed with 300 mL of hot deionized water. This process was repeated four times to ensure the removal of excess ammonium salt of diamine. The final product obtained by filtration was dried under vacuum at 65°C for 6 h.

The PI/Clay nanocomposite membranes were prepared by mixing suitable amounts of the PI and clays in a flask for a particular concentration. The reaction mixture was agitated to high speed stirring at 80°C for 4 h and at 60°C for 20 h for uniform dispersion of organoclay platelets in the PI matrix. Various compositions (1, 3 and 5%) of organoclays were prepared by mixing various amounts of clays to the PI solution (Scheme 3). Thin PI/Clay films of uniform thickness were obtained by pouring the hybrid solutions into Petri dishes, followed by solvent evaporation at high temperature. These films were further dried at 85°C under reduced pressure to a constant mass.

Scheme 3.

Fabrication of organoclay and PI/Clay nanocomposites.

Results and discussion

FT-IR spectra of nanocomposite membranes

Figure 4 shows the FT-IR spectra of the (a) Na-Clay, (b) organoclay, (c) pure PI and (d) PI/Clay 1% materials. FT-IR spectra of Na-Clay and organoclay are shown in Figure 4(a) and (b). The FT-IR spectrum of Na-Clay showed the characteristic absorption bands of the Si–O, Al–O and Mg–O moieties at 1039, 513 and 462 cm⁻¹, respectively. The incorporation of organic groups in organoclay was confirmed by the presence of peaks at 3050 cm⁻¹ (ν C = C–H), 2850 cm⁻¹ (νNH₂⁺) and 1650 cm⁻¹ (δC = C aromatic). In the FT-IR spectrum of PAA, the characteristic absorption bands of amides and carboxyl groups appeared at 3335 and 1665 cm⁻¹, respectively, whereas they disappeared in the spectrum of the imide.²⁷ The appearance of new peaks at 1782 cm⁻¹ (asymmetric C = O stretching), 1372 cm⁻¹ (νC–N) and 1050 and 742 cm⁻¹ (ring deformation) indicated the complete conversion of PAA to PI. In the nanocomposite sample, the presence of a peak at 1050 cm⁻¹ corresponding to the Si–O stretch indicated the incorporation of organoclay in the PI matrix (Figure 4(d)).

Figure 4.

FT-IR spectra of the (a) Na-Clay, (b) organoclay, (c) pure PI and (d) PI/Clay 1% materials.

XRD patterns of nanocomposite membranes

Figures 5(a)–(f) shows the XRD patterns of clay samples and polymer nanocomposite samples, respectively. For Na-Clay, the diffraction pattern shows a distinct peak at 2θ = 4.68° corresponding to 18.22 A°, which is assigned to the 001 lattice spacing of montmorillonite. After modification with dodecyltrimethylammonium

Figure 5.

XRD patterns of the (a) Na-Clay, (b) organoclay, (c) pure PI and (d) PI/Clay 1%, 3% and 5%materials.

diamine, 2θ decreased to 3.83°, i.e. the d spacing increased to 21.25°A for organoclay. The greater spacing in the organoclay, when compared to Na-Clay, would help the silicate layers to disperse easily.

In nanocomposite samples, no distinct peak was observed at lower clay concentrations (up to 3%), whereas for 5% clay concentrations, a distinct peak at about 2θ = 3.98° (d = 14.4 A°) was observed. Usually, there are two types of nanocomposites depending upon the dispersion of clay particles. The first type is an intercalated polymer–clay nanocomposite, which consists of well-ordered multilayers of polymer and silicate layers a few nanometers thick and these structures would give rise to distinct peaks in XRD patterns as in the case of 5% nanocomposite sample. In these nanocomposite samples, the extent of polymer penetration was not sufficient to delaminate the ordered structures. The second type is an exfoliated polymer–clay nanocomposite, in which is there is a loss of ordered structures due to the extensive penetration of polymer into the silicate layers. Such types would not produce distinct peaks in the XRD patterns. As there was no coherent XRD signal at lower clay concentration (up to 3%), one can say that these nanocomposite samples were exfoliated or delaminated.²⁸

Morphology of nanocomposite membranes

Figures 6(a) and (b) presents the TEM images of PI/Clay 3% and PI/Clay 5% nanocomposites. In these images, part of clay nanoplatelets denoted as the dark lump have been intercalated and some individual nanoplatelets disperse homogeneously in PI matrix with a slice size lower than 0.5 μm. A minor extent of clay reveals a number of agglomerates in PI/Clay 3%; however, the dispersion of individual silicate layers in the polymer matrix for PI/Clay 5% is less homogeneous than PI/Clay 3%. In Figure 6(c), the magnified sample edge of PI/Clay 5% is given to present the morphology of the nanocomposites more clearly. From Figure 6(c), we confirm that the PI/Clay nanocomposites are prepared successfully and the nanoplatelets are embedding in the PI matrix as we think. It also can be seen the distribution of the clays in the PI matrix which depends on the amount of the clay which is consist with the result of XRD.²⁹ Even so, both PI/Clay 1% and PI/Clay 5% are considered as a partial intercalated system.

Figure 6.

The TEM micrographs of (a) PI/Clay 3%, (b) PI/Clay 5% and (c) The sample edge of PI/Clay 5%.

The FE-SEM micrograph of pure clay is shown in Figure 7(a), while neat PI and the nanocomposites FE-SEM images are displayed in Figure 7(b)–(d). The image of clay reveals an apparent layered structure and the original stack size of lamellar structure is close to 3–5 μm. From the cross-section images of the PI/Clay composites, it can be seen that the surface morphology changes according to the clay loading. For brevity, we just list some representative micrographs. Neat PI shows a smooth surface, and the evenly dispersed white dots in the PI matrix are residual Zinc catalysts which have been verified by EDAX (Energy-dispersive X-ray spectroscopy). Moreover, PI/Clay 3% reveals a homogenous dispersion of clay in the polymer matrix, though the surface becomes a little rough. The image of PI/Clay 5% displays a rougher surface and presents some microscale fillers and gaps due to the agglomeration of individual silicate layers. Both the micrographs and XRD results indicate that the organic modified clays (organoclay) can homogeneously disperse in the polymer matrix at a small amount; however, agglomeration may form with higher loadings of clay added.

Figure 7.

The FE-SEM micrographs of (a) Pure Clay, (b) Pure PI (c) PI/Clay 3% (d) PI/Clay 5%.

TGA Analysis of PI/Clay nanocomposite membranes

In order to investigate the thermal stability of the PI/Clay nanocomposites, thermogravimetric analysis (TGA) was performed. The TGA was performed under nitrogen atmospheres. The TGA weight loss curves and first derivative TGA curves for reference PI and the nanocomposite samples under nitrogen atmosphere (pyrolytic environment) are presented in Figure 8. In nitrogen atmosphere the pure PI and the PI/Clay nanocomposites exhibited only one step decomposition. The main degradation process during TGA analysis under nitrogen atmosphere is due to the combined degradation processes—polymer chain degradation through end group initiated mechanism and the thermal degradation of the products formed during polymer chain degradation. Under nitrogen environment, the thermal decomposition of polymer is leading to thermally stable cross-lined carbonaceous species which are not undergoing any further decomposition due to the presence of inert nitrogen atmosphere and hence the large amounts of reside at the end of TGA analysis. The results of TGA analysis such as temperature at 1% weight loss, temperature at 5% weight loss, onset of decomposition, and the amount of residue at end of TGA analysis (at 750°C) are summarized in the Table 1. It is clear from the results that the temperatures at 1% and 5% weight loss as well as the onset of decomposition is lower for the nanocomposites compared to pure PI. This observation could be due to the degradation of diamine modifier which is present in Na-Clay. The diamine present in the Na-Clay is known to undergo Hoffman degradation around 190°C.^30,31 The amount of residue left over on TGA analysis (at 750°C) under nitrogen atmosphere for PI/Clay nanocomposites is higher than that of the pure PI.

Figure 8.

(a) TGA weight loss curves and (b) first derivative TGA curves for PI/clay nanocomposites during TGA analysis under nitrogen atmosphere.

Table 1.

The results of TGA analysis on pure PI and P/Clay membranes.

Material	Temp. At 1% weight loss	Temp. At 5% weight loss	Onset of decomposition	Residual mass at 750°C
Pure PI	401.27 ± 22.51	425.26 ± 13.61	445.88 ± 12.35	41.58 ± 1.25
PI/Clay 1%	399.21 ± 32.41	423.87 ± 21.36	443.21 ± 25.37	42.19 ± 1.89
PI/Clay 3%	398.14 ± 56.32	421.89 ± 12.32	442.32 ± 18.20	47.26 ± 2.36
PI/Clay 5%	396.23 ± 15.31	421.12 ± 10.25	440.41 ± 12.06	50.01 ± 4.22

DSC Analysis of PI/Clay nanocomposite membranes

The influence of clay on thermal properties and crystallization behaviour of PI was investigated by using differential scanning calorimetry (DSC). Thermal properties of PI/Clay membranes, such as glass transition temperature (T_g) crystallization temperatures (T_c) and melting temperatures (T_m) were obtained from the DSC analysis. The results of the DSC analysis are presented in the Table 2. It is clear from the results in the Table 2 that incorporation of clay into the PI matrix reduces the T_g of the PI matrix. The reason for the reduction in T_g with incorporation of clay could be due to the higher polymer chain mobility in presence of clay. In addition, the modifier (diamine) present in the clay could act as a plasticizer and increase polymer chain motion. Similar reduction in T_g is reported in the literature for nanocomposites based on PI and organo-modified clay. Addition of clay platelets into PI has very little influence on melting temperature, T_m2 of nanocomposite samples. However, incorporation of clay into PI slightly increases the crystallization temperature, T_c of nanocomposites. With increase in clay loading, the T_c of nanocomposites gradually increases. In addition, the degree of crystallinity of the PI/Clay nanocomposites slightly increased with increase in clay platelets content. This increase in crystallization temperature and the degree of crystallinity in the nanocomposites could be due to the nucleation effect of clay platelets in PI matrix. Presence of nano-reinforcements, such as clays and CNTs are known to exhibit nucleating effect in polymer matrices.

Table 2.

Results of DSC analysis of PI and PI/Clay nanocomposites T_g, T_m1, T_m2 and T_c.

Material	T_g	T_m1	T_m2	T_c
Pure PI	285.14 ± 19.45	568.36 ± 25.36	577.58 ± 10.23	444.87 ± 24.99
PI/Clay 1%	284.89 ± 10.58	569.24 ± 12.72	578.45 ± 25.16	446.32 ± 17.30
PI/Clay 3%	282.47 ± 21.85	569.89 ± 19.02	579.88 ± 21.66	447.17 ± 14.88
PI/Clay 5%	282.56 ± 19.35	570.21 ± 52.12	580.63 ± 20.55	449.99 ± 9.25

A double melting peak appeared for both pure PI and the PI/Clay nanocomposites. This behavior could be due to the presence of a dual lamella thickness distribution which is formed during crystallization of PI. The nanocomposite samples also presented double melting peaks, indicating that added clay has no influence on the crystal morphology in the PI/Clay nanocomposites. As seen from the thermograms, the position of the first melting (T_m1) is unaltered and the second melting peak temperature (T_m2) is slightly affected by presence of clay filler. The low temperature first melting peak is observed at about 568–570°C while the high temperature main peak position varies between 577–580°C depending on the clay content. In addition, with increase in clay loading, the first melting peak shoulder is less pronounced, being obvious for the nanocomposite containing 5 wt% of clay. This is probably because clay at higher loadings starts to influence crystallization behaviour and morphology. The DSC crystallization endotherms of pure PI and PI/Clay nanocomposites are presented in Table 2 and Figure 9. A slight shift in the crystallization peak position towards the higher temperature is observed for the nanocomposites containing clay. This means that the nanocomposite samples crystallize at higher temperature compared to the pure PI, as also seen from the corresponding T_c values in Table 2. The shift in crystallization peak towards higher temperature is due to the heterogenous nucleating effect of dispersed clay mineral in PI matrix.

Figure 9.

DSC thermographs of pure PI and PI/Clay nanocomposites.

Mechanical properties of PI/Clay nanocomposite membranes

The tensile properties of pure PI and PI/Clay nanocomposites with various clay contents are reported in Figure 10. Compared to pure PI, the tensile strength, elongation at break and modulus of PI/Clay nanocomposites dramatically increased with increasing clay content up to 3 wt%. This is due to the exfoliated structure of silicate layer providing the optimum reinforcement. However, the tensile properties decreased when a high content (5 wt%) of clay platelets was added. This suggests that the clay partially remains the intercalated and stacked structure, causing the weakly reinforcing effect, the non-idealized nanoclay dispersion, and the presence of microvoids as shown in FE-SEM micrographs.

Figure 10.

Tensile strength, Elongation at break and Modulus of pure PI and PI/Clay nanocomposites with various inorganic nanoclay contents.

Tensile strength of the PI/Clay nanocomposites was highly dependent upon the clay platelets concentrations and was improved with increased clay concentration. This is due to the high strength of the clay platelet. Initially, the strength was improved by 12% for only 1 wt% clay incorporation. At higher concentrations, the rate of modulus increase became less, e.g., there is only 15% increase of the strength for 3 wt% clay by sonication, which can be attributed to the presence of unexfoliated aggregates in the structure under higher clay concentrations. The strength was also related to the method of clay dispersion; nanocomposites by sonication had slightly higher strength than those by shear mixing. Better dispersions by sonication increased the interfacial contact area between the clay particles and PI matrix, thus providing a better coupling and a more efficient restriction of epoxy chain mobility under load.³²

Tensile modulus was also related to the clay dispersion. Sonication dispersed PI/Clay nanocomposites lead to improved tensile modulus, and an increase of 25% in tensile modulus was observed for only 1 wt% clay platelets content. The nanocomposite modulus at 5 wt% clay platelets concentration lead to decrease and it is still near that of neat PI. Overall, exfoliated structures lead to improved modulus and strength, while the intercalated structures lead to decreased tensile strength and modulus. With sonication, exfoliation and intercalation structures probably coexisted at higher clay concentration with intercalated flocs as defects thus resulting in reduced mechanical strength. The interfacial interaction between clay and matrix is thus reduced by poorer dispersion at high clay concentration and crack fracture could initiate in clay galleries. Decreased tensile and modulus strength has also been attributed to the presence of aggregation of clay particles which is due to the inefficient dispersion or clay concentrations above the percolation threshold. Aggregation not only reduces the interfacial interactions between clay and polymer but also generate macrocracks. Large aggregates and intercalated structures co-existed in shear mixed composites, so there was even less interfacial interaction between the clay and the PI matrix compared to the sonication dispersed composites, as will be observed in FE-SEM and TEM images.³³

Gas permeability studies of PI/Clay nanocomposite membranes

Gas permeation of neat PI and PI/Clay nanocomposites were assessed in terms of gas permeability and selectivity for gas pairs (O₂/N₂ and CO₂/N₂). The permeance and the selectivity data were depicted in Figure 11 and Figure 12, respectively. The permeation of CO₂, N₂, and O₂ gases was increased with the increase in the loading percentage of clay. Besides, the selectivity of O₂/N₂ and CO₂/N₂ gas pairs were decreased with an increase in clay loading percentage. The relative increase in permeance for CO₂ from pure PI to PI/Clay nanocomposite (3 wt%), was about 245% from 17.98 to 61.25 GPU, while the CO₂/N₂ selectivity was decreased by 17.21% from 26.01 to 21.05. The corresponding increase in permeance for O₂ was 240% from 4.91 to 16.89 GPU with a drop of 7% in O₂/N₂ selectivity from 5.76 to 5.21.

Figure 11.

Permeance of pure PI and PI/Clay nanocomposites.

Figure 12.

Selectivity of pure PI and PI/Clay nanocomposites for CO₂/N₂, and O₂/N₂ gas separation.

In both pure PI and PI/Clay nanocomposites, the order of gas permeability was as follows: P (CO₂) > P (O₂) > P (N₂), was in consistent with most polymeric gas separation membranes. Gas permeability was decreased with the kinetic diameter and increased with condensability of the penetrant gas was reliable with glassy polymers. Kinetic diameter and gas condensability data were presented in Table 3 explains the gas permeability order.^34,35 The increase in gas permeation rate is attributed to the higher permeability via the pillared clay present within PI/Clay nanocomposites. Clay contributes increased in gas permeability due to the rise in micropore and total pore volume reported by Timofeeva et al.³⁶ As depicted in Figure 11, the permeation rate improvement for gases was similar up to clay of 3 wt% loading, whereas the permeation increased significantly for PI/Clay nanocomposites (5 wt%), especially for slow permeable N₂ gas. This can be ascribed with the development of unselective voids at the PI and clay platelets interface when the clay loading increased to 5 wt%.²³ It was confirmed by the cross‒sectional morphology of the PI/Clay nanocomposites. The CO₂/N₂ and O₂/N₂ gas pair selectivities for PI/Clay nanocomposites (1 wt%) and PI/Clay nanocomposites (3 wt%) were similar to the pure membrane. (Figure 12). Besides, the selectivity of O₂/N₂ and CO₂/N₂ gas pairs were decreased with an increase in organoclay loading percentage to 5 wt% owing to better enhancement in permeance of slow (N₂) gas than fast (O₂) gas permeation rate. The PI/Clay nanocomposites (5 wt%) shown higher permeability with the lower selectivities for CO₂/N₂ and O₂/N₂ gas pairs compare to other fabricated membranes. The ability of PI/Clay nanocomposites (5 wt%) for discerning the permeance gas based on the kinematic diameter of gas molecule was diminished as the organoclay loading increased within the membrane matrix. Thus, the PI/Clay nanocomposites (5 wt%) might permit the slow permeable (N2) gas despite of gas kinematic diameter. This leads to higher permeance of N₂ gas via the PI/Clay nanocomposites (5 wt%) which results in decreased selectivites for CO₂/N₂ and O₂/N₂ gas pair. PI/Clay nanocomposites (3 wt%) exhibit better selectivity for gas pairs with reasonable permeability. Thus, the optimal loading of organoclay within the fabricated PI/Clay nanocomposites was found to be at 3 wt% loading. PI/Clay nanocomposites (3 wt%) exhibit a significant increase in permeability, fractional decrease in selectivity, and driving the performance nearer to the Robeson upper bound. High filler loading leads to compromise of the structural integrity and possible filler agglomerations in the membrane microporous region.²⁴ Hence organoclay competes against other fillers even at a lower loading percentage and exhibits comparable results to the other fillers.^24,37,38

Table 3.

Kinetic diameter and gas condensability of studied gases from literature.³⁴

Gas	Kinetic diameter (Å)	Condensability (K)
CO₂	3.30	195
N₂	3.64	71
O₂	3.46	107

Comparison of fabricated nanocomposites with literature

The fabricated nanocomposites were compared with literature to ascertain credibility and conformity. The properties of the fabricated nanocomposites are in concordance with published pre‒existing data denoted in Table 4. The permeability of the membranes was estimated by assuming an active skin layer thickness of 150 nm as commonly reported active layer thickness for membranes.²³ The permeance and selectivity were compared with MCM‒41/PSf membranes studied by Jomekian et al.,³⁹ and Clay/PSf membranes investigated by Zulhairun et al.²³ in Table 4. PI/Clay nanocomposites were compared with other PI membranes incorporated with other filler particles such as silica, and carbon nano‒fibers reported by Ahn et al.,⁴⁰ and Jomekian et al.⁴¹ The fabricated PI/Clay nanocomposites tend to show improved selectivity and permeability. The beneficial aspect was the enhanced properties exhibited with a mere 1% loading of Clay as compared to 20% silica loading. The requirement of low loading Clay can be attributed to the excellent compatibility between the pillared clay and PI when compared to the higher loading fractions of inorganic silica,⁴⁰ and MCM filler.^39,41

Table 4.

Comparison of O₂ and N₂ permeability and O₂/N₂ selectivity of fabricated MMMs with the reported values in literature with the reported values in literature.

Membrane	P_O2	P_N2	α_O2/N2	Pressure (bar)	Reference
U MCM‒41 (20 wt%)/PSF	5.35	1.00	5.35	4	^41,43
PSF –Clay (1.0 wt%)	16.37	3.19	5.13	5	^{37, 42 and 44}
PI/Clay 1%	18.52	3.45	5.32	4	This study

Conclusion

Polyimide containing bis three flours groups designated was synthesized. Then PI/Clay nanocomposites with different concentration of clay platelets were fabricated. The gas permeation, thermal, mechanical and morphological properties were investigated in the nanocomposites. The results of XRD, TEM, and FE-SEM indicated that clay could well disperse in the PI matrix and were intercalated by diamine and PI macromolecules. The nanocomposites a slightly higher crystallisation temperature compared to pure PI. The increased crystallization temperature of nanocomposites is due to a heterogenous nucleating effect of clay. The thermal stability and glass transition temperature of nanocomposite at lower clay loadings are similar to PI but gradually decrease with increase in clay content. The tensile properties of PI/Clay nanocomposites showed the highest values at the optimum clay content of 3 wt%. The experimental observations indicate a significant increase in membrane permeability at a low loading of clay with an insignificant drop in permselectivity. The most efficient membrane was found to be at 3 wt % loading of clay. Organoclays are a relatively novel filler medium to be used in nanocomposites, and exhibit favorable compatible properties for gas separation. The study shows that there is a scope and potential application for the implementation of montmorillonite silicates for gas separation.

Footnotes

Acknowledgements

H.Ahmadizadegan. acknowledges financial support from Iran Nanotechnology Initiative Council (INIC).

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Correction (July 2023):

This article has been updated with branch name in the affiliations since its original publication.

ORCID iD

Hashem Ahmadizadegan

References

Chung

Jiang

, et al. Mixed matrix membranes (MMMs) comprising organic polymers with dispersed inorganic fillers for gas separation. Progress in Polymer Science 2007; 32: 483–507.

Te Hennepe

HJC

Boswerger

WBF

Bargeman

, et al. Zeolite-filled silicone rubber membranes – experimental determination of concentration profiles. Journal of Membrane Science 1994; 89: 185–196.

Te Hennepe

HJC

Bargeman

Mulder

MHV

, et al. Permeation through zeolite filled silicone rubber membranes. Studies in Surface Science and Catalysis 1998; 39: 411–420.

Zulhairun

Ismail

Matsuura

, et al. Asymmetric mixed matrix membrane incorporating organically modified clay particle for gas separation. Chem Eng J 2014; 241: 495–503.

Urban

. Hydrogen-bonded polyimide/metal-organic framework hybrid membranes for ultrafast separations of multiple gas pairs. Adv Funct Mater 2019; 29: 1903243.

Wang

Qin

, et al. Interfacial engineering in metal–organic framework-based mixed matrix membranes using covalently grafted polyimide brushes. J Am Chem Soc 2018; 140: 17203–17210.

Ahmadizadegan

Tahriri

, et al. Polyimide-TiO2 Nanocomposites and their corresponding membranes: synthesis, characterization, and gas separation applications. Solid State Sci 2019; 89: 25–36.

Ahmadizadegan

Ranjbar

Esmaielzadeh

. Synthesis and characterization of green membranes polyimide/titania bionanocomposites containing amino acid and benzimidazole moieties for gas transport properties. Polym Eng Sci 2017; 58: 1666–1677.

Mallakpour

Ahmadizadegan

. Manufacture of zinc oxide/chiral poly(amide-imide)-functionalized amino acid and thiazole bionanocomposites: using ionic liquid and ultrasonic irradiation. Journal of Thermoplastic Composite Materials 2015; 28: 672–685.

10.

Chang

J-H

. Polyimide nanocomposites with functionalized graphene sheets: thermal property, morphology, gas permeation, and electroconductivity. Journal of Thermoplastic Composite Materials 2018; 31: 837–861.

11.

Pandey

Dayanidhi

Mohanty

, et al. Effect of clay loading on flammability of poly(methyl methacrylate)/clay nanocomposites. Journal of Thermoplastic Composite Materials 2013; 26: 663–679.

12.

Zaman

Hun

Khan

, et al. Polypropylene/clay nanocomposites: effect of compatibilizers on the morphology, mechanical properties and crystallization behaviors. Journal of Thermoplastic Composite Materials 2014; 27(3): 338–349.

13.

Bousmina

. Study of Intercalation and exfoliation processes in polymer nanocomposites. Macromolecules 2006;39(12):4259–4263.

14.

Fornes

Hunter

Paul

. Nylon-6 nanocomposites from alkylammonium-modified clay: the role of alkyl tails on exfoliation. Macromolecules 2004;37(5):1793–1798.

15.

Zhao

Tang

Qin

, et al. Effects of surfactant loadings on the dispersion of clays in maleated polypropylene. Langmuir 2003; 19(18): 7157–7159.

16.

Meng

Wang

, et al. The investigation of exfoliation process of organic modified montmorillonite in thermoplastic polyurethane with different molecular weights. Compos Sci Technol 2008; 68: 1815–1821.

17.

Balazs

Singh

Zhulina

. Modeling the interactions between polymers and clay surfaces through self-consistent field theory. Macromolecules 1998; 31: 8370–8381.

18.

Balazs

Singh

Zhulina

, et al. Modeling the phase behavior of polymer/clay nanocomposites. Acc Chem Res 1999; 32: 651–657.

19.

Fermeglia

Ferrone

Pricl

. Computer simulation of nylon-6/organoclay nanocomposites: prediction of the binding energy. Fluid Phase Equilib 2003; 212: 315–329.

20.

Tanaka

Goettler

. Predicting the binding energy for nylon 6,6/clay nanocomposites by molecular modeling. Polymer 2002; 43: 541–553.

21.

Toth

Coslanich

Ferrone

, et al. Computer simulation of polypropylene/organoclay nanocomposites: characterization of atomic scale structure and prediction of binding energy. Polymer 2004; 45: 8075–8083.

22.

Meng

Wang

Zhao

, et al. Morphology evolutions of organically modified montmorillonite/polyamide 12 nanocomposites. Polymer 2007; 48: 2508–2519.

23.

Zulhairun

Ismail

. The role of layered silicate loadings and their dispersion states on the gas separation performance of mixed matrix membrane. J Memb Sci 2014; 468: 20–30.

24.

Moore

Koros

. Non-ideal effects in organic-inorganic materials for gas separation membranes. J Mol Struct 2005; 739: 87–98.

25.

Ahmadizadegan

Ghavvas

Ranjbar

, et al. Synthesis and characterization of fluorinated polyimide/TiO2 nanocomposites: enhancement of separation of four gases, thermal, optical and mechanical properties. Polym Bull 2018; 75: 2729–2750.

26.

Mallakpour

Dinari

. Novel nanocomposites based on reactive organoclay of l-tyrosine and amine end-capped poly(amide–imide): synthesis and characterization. Appl Clay Sci 2013; 75: 67–73.

27.

Gopala Krishnan

Wisanto

Osiyemi

, et al. Synthesis and properties of BCDA-based polyimide–clay nanocomposites. Polym Int 2007; 56: 787–795.

28.

Chen

Tien

Wei

. Synthesis and characterization of novel segmented polyurethane/clay nanocomposites. Polymer 2000; 41: 1345–1353.

29.

Amin

Samy

, Preparation and characterization of some hyperbranched polyesteramides/montmorillonite nanocomposites, International Journal of Polymer Science 2013; 2013: 1–8.

30.

Chowreddy

Nord-Varhaug

Rapp

. Recycled Poly(Ethylene Terephthalate)/Clay nanocomposites: rheology, thermal and mechanical properties. J Polym Environ 2019; 27: 37–49.

31.

Ray

Bousmina

Okamoto

. Structure and properties of nanocomposites based on poly (butylene succinate-co-adipate) and organically modified montmorillonite. Macromol Mater Eng 2005; 290: 759–768.

32.

Luecha

Magaraphan

. Thermal and mechanical properties of Nanoclay-Poly(acrylic acid) gel reinforced polyethylene Terephthalate/Clay Nanocomposite. E-J Surf Sci Nanotech 2015; 13: 107–110.

33.

Pegoretti

Kolarik

Peroni

, et al. Recycled poly(ethylene terephthalate)/layered silicate nanocomposites: morphology and tensile mechanical properties. Polymer 2004; 45: 2751–2759.

34.

Sadeghi

Semsarzadeh

Barikani

, et al. The effect of urethane and urea content on the gas permeation properties of poly (urethane-urea) membranes. J Memb Sci 2010; 354: 40–47.

35.

Natarajan

Sasikumar

Elakkiya

, et al. Pillared cloisite 15A as an enhancement filler in polysulfone mixed matrix membranes for CO2/N2 and O2/N2 gas separation. J Nat Gas Sci Eng 2021; 86: 103720.

36.

Timofeeva

Khankhasaeva

Badmaeva

, et al. Synthesis, characterization and catalytic application for wet oxidation of phenol of ironcontaining clays. Applied Catalysis B: Environmental 2005; 59: 243–248.

37.

Ahmadizadegan

Esmaielzadeh

. The role of organically modified clay particle on thermal, mechanical and gas barrier properties of polyimide nanocomposites and toward improvement of gas selectivities. Polym Technol Mater 2020; 59: 1855–1874.

38.

Ahmadizadegan

Esmaielzadeh

. Preparation and application of novel bionanocomposite green membranes for gas separation. Polym. Bull 2019; 76: 4903–4927.

39.

Jomekian

Pakizeh

Shafiee

, et al. Fabrication or preparation and characterization of new modified MCM-41/PSf nanocomposite membrane coated by PDMS. Sep Purif Technol 2011; 80: 556–565.

40.

Ahn

Chung

Pinnau

, et al. Polysulfone/silica nanoparticle mixed-matrix membranes for gas separation. J Membr Sci 2008; 314: 123–133.

41.

Jomekian

Mansoori

SAA

Monirimanesh

, et al. Gas transport behavior of DMDCS modified MCM-48/polysulfone mixed matrix membrane coated by PDMS. Korean J Chem Eng 2011; 28: 2069–2075.

42.

Manias

Chen

Krishnamoorti

, et al. Intercalation kinetics of long polymers in 2 nm confinements. Macromolecules 2000; 33: 7955–7966.

43.

Zhang

, et al. Dynamics of polymers in organosilicate nanocomposites. Macromolecules 2003; 36: 823–829.

44.

Kadar

Szazdi

Fekete

, et al. Surface Characteristics of layered silicates: influence on the properties of clay/polymer nanocomposites. Langmuir 2006; 22(18): 7848–7854.