Highly selective thin film composite hollow fiber membranes for mixed vapor/gas separation

Abstract

In the present study, thin film composite membranes have been prepared using an interfacial polymerization method. First, we coated a polydopamine (PDA) layer using different concentrations of PDA solution on polyethersulfone (PES) hollow fiber supports. After the PDA coating layer, thin film composite (TFC) membranes were prepared with 3,5-diaminobenzoic acid (3,5-DABA) as an aqueous phase monomer and trimesoyl chloride (TMC) as an organic phase monomer to synthesize a hydrophilic polyamide layer. This prepared selective layer is considered desirable to fabricate hydrophilic TFC membranes for water vapor/N₂ separation. The TFC membranes by interfacial polymerization were confirmed and discussed using the accumulated results of characterization. Hollow fiber membranes (HFM) surface modification with PDA before TFC coatings was proposed to modestly and effectively enhance the membrane selectivity. The newly prepared TFC hollow fiber membranes acquired reasonably excellent selectivity and superior permeation fluxes. As a result, membrane sample MS4 coated with 2.0 wt% PDA and TFC prepared using 0.5 wt% of 3,5-DABA with 0.2 wt% of TMC and 60 s of reaction time showed the best permeance and selectivity as 3185 GPU and 195, respectively, compared with other TFC membranes prepared with different PDA concentrations. Overall, the membranes showed good performance in the entire range of operating conditions investigated.

---

1. Introduction

The removal of water vapor from gas streams is an important industrial process. It is predicted that more than 20% of the population in the world will suffer from water shortage in 2025 mainly due to industrialization, climate change and accelerated desertification.¹ Representatively, the largest amount of water vapor (∼250 tons per h) discharges are from coal power plants of power 500 MW.² In the U.S., 39% brackish water is normally utilized for power generation, whereas steam-power plants consumption is about 71% water. Flue gases generated and emitted from the combustion process generally contain 6–18% volume of water vapor.^3,4 Most of the consumed water resources are discharged in the vapor phase. There are several conventional technologies applied to remove water vapor from gas streams. For example, water vapor from a gas stream can be removed using a condenser, liquid adsorption,^5–9 a desiccant drying system^10,11 or a membrane system.^12–14 The condensation process has some disadvantages such as condensed water is relatively dirty and corrosive; whereas a desiccant system, which is a widely used and accepted system, requires regeneration of the desiccant, which is an energy-intensive process.

However, membrane technology is an attractive, energy-efficient alternative for molecular separations because of its superior advantages over other processes. The membrane process has high energy efficiency (no phase change is required to achieve separation), reliability (no moving parts), and small footprint. It allows the selective removal of water vapor from (flue) gas streams and can produce water with high purity without additional heating. These advantages make membrane technology an interesting and promising alternative for removing water vapor from flue gasses in coal-fired power plants.¹⁵

In this study, vapor permeation was employed for the dehydration of flue gas. Vapor permeation is similar in principal to pervaporation, but systematically constitutes a mixture of vapor or gases.¹⁶ Vapor permeation takes place through the membrane by a solution–diffusion mechanism. Both molecular interactions between the membrane and the separated species¹⁷ and materials are determining factors for membrane separation, which evaluate the membrane performance. In relation to transport of water vapor, dehydration of flue gases,^5,18 drying of natural gas¹⁹ and drying of compressed air,²⁰ the polymers used are selective as membrane materials. If the water vapor recovered retains a high quality through a polymer, which induces high water vapor selectivity, the obtained permeants can be reused for water supply in the steam cycle process.²¹ Polyethersulfone (PES) is an established membrane material and commonly used in the manufacture of various membrane systems.

Herein, a polymeric, asymmetric and composite membrane was used for water vapor/N₂ mixed gas separation. The coating of polydopamine (PDA) provides hydrophilicity to the hydrophobic membrane surface. PDA can be firmly attached to most inorganic and organic surfaces and maintains stability of the coating area even on wet surfaces similar to adhesive proteins inspired by mussels.²² 3,5-Diaminobenzoic acid (3,5-DABA) containing both amine and carboxylic acid groups was used as an aqueous monomer and reacted with trimesoyl chloride (TMC) to make a selective thin film composite membrane. Especially, 3,5-DABA is applied for polymeric gas separation membranes; it is considered to be a way to select the aqueous phase material and focus on the effects of preparation conditions, such as concentration of PDA, which plays an important role in the performance of polymeric membranes. In this study, the performance of the membranes fabricated using an interfacial polymerization technique was successfully examined for water vapor/N₂ mixed gas.

2. Experimental

2.1 Membrane preparation

2.1.1. Materials. Polyethersulfone (PES, Ultrason® E6020P, BASF, Germany) and lithium chloride (LiCl, Sigma-Aldrich) as a pore former were used as received. N,N-Dimethylformamide (DMF, SAMCHUN chemicals) was used as a solvent to prepare the hollow fiber support layer. Dopamine (DA), to increase the hydrophilicity of the hollow fiber membranes, 3,5-diaminobenzoic acid (3,5-DABA) and trimesoyl chloride (TMC) were purchased from Sigma-Aldrich as monomers for interfacial polymerization. n-Hexane (99.9%, Fisher Scientific, NJ) was used as a solvent for TMC. For the water vapor/N₂ mixture gas, feed gas of pure nitrogen (99.9%) and deionized water (DI) from a Milli-Q ultrapure water purification system (Millipore) was used. All chemical reagents for this study were used without further purification.

2.1.2. Preparation of polyethersulfone hollow fiber (HF) support membranes. The hollow fiber membranes were prepared using a phase inversion technique similar to methods discussed elsewhere.^23,24 ESI Fig. S1† represents the schematic of the hollow fiber membrane spinning system. A dry jet-wet spinning process was used for the fabrication of the hollow fiber membranes. After spinning, the fabricated membranes were left in water for a few days to remove residual solvent and then dried at ambient temperature by hanging them vertically for 1–2 days. In this present study, the dope solution and internal coagulant (D.I. water) were passed through a double pipe spinneret having 0.16/0.9 mm inner/outer diameter and an air gap maintained at 0.5 cm. The dope solution composed of 18.0 wt% PES and 77.0 wt% N-methylpyrrolidone as the solvent and 5.0 wt% lithium chloride as the additive. Detailed conditions for preparation of the hollow fiber membrane are listed in Table 1.

Table 1 Composition of dope solution and spinning conditions for the preparation of PES hollow fiber membrane

Composition
PES	18.0 wt%
NMP	77.0 wt%
LiCl	5.0 wt%

---

Spinning conditions
Internal coagulant	Distilled water
Injection rate of dope solution	5.5 mL min^−1
Injection rate of internal coagulant	2.5 mL min^−1
Winding speed	18 m min^−1

---

2.1.3. PDA coating and preparation of thin film composite (TFC) membranes. The TFC membrane was prepared using different concentrations of PDA coated HFM with 3,5-diaminobenzoic acid and TMC to synthesize a more hydrophilic polyamide/polyester layer, as shown in Fig. 1 (PDA self-polymerization) and Fig. 2 (TFC cross linked polyamide/polyester structure). For the fabrication of the hydrophilic-supported TFC membrane, porous PES hollow fibers were immersed in 2.0 g L⁻¹ DA-HCl at pH 8.5 (PBS buffer solution) before interfacial polymerization for 30 min. PDA has a tendency to form free PDA particles at a high dopamine concentration. The polydopamine coating is very easy to accomplish and does not need any catalyst, organic solvent or rigorous reaction conditions. To synthesize the cross-linked selective layer on PDA-coated HF membranes by interfacial polymerization, 3,5-DABA and TMC were used as the aqueous and organic phase monomers, respectively. 3,5-DABA was introduced as the aqueous phase monomer to fabricate TFC membranes because it has the hydrophilic group, i.e. carboxylic acid, and the TMC monomer has high reactivity with the other monomer and is commonly used in the organic phase. The PES HF membranes were dried in advance and immersed in an aqueous solution of 3,5-DABA for 5 min followed by draining off for 2–5 min to remove excess solution. They were then immersed in 0.2 wt% of TMC in hexane solution for 60 seconds reaction time, followed by draining off of the excess solution. The process flow chart of TFC preparation using the interfacial polymerisation method is shown in Fig. 3. The conditions for the preparation of all the TFC membranes are mention in Table 2.

Fig. 1 Self-polymerization of dopamine after oxidation.

Fig. 2 Crosslink structure after polymerization by DABA and TMC forming the polyamide thin film selective layer.

Fig. 3 Process flow chart of TFC preparation by interfacial polymerization.

Table 2 The conditions for the preparation of TFC membranes; the concentrations of PDA, 3,5-DABA and TMC are weight percentages (wt%)

Entry	PDA (wt%)	3,5-DABA (wt%)	TMC (wt%)	Reaction time (s)
MS1	0.25	0.5	0.2	60
MS2	0.50	0.5	0.2	60
MS3	1.0	0.5	0.2	60
MS4	2.0	0.5	0.2	60

---

2.2 Characterization of TFC membranes

2.2.1. ATR-FTIR analysis. ATR-FTIR analysis technique is a useful and desirable technique for membrane analysis. Verification for the presence of certain functional groups, attributed to the interfacial polymerization process on the sub-layer surfaces, was conducted by FTIR for surface chemistry of real specimens. ATR-FTIR spectra were obtained using an ALPHA-P Spectrometer (Bruker Optic GmbH) with diamond crystal in the range of 600–4000 cm⁻¹.

2.2.2. X-ray photoelectron spectroscopy (XPS) analysis. XPS analysis was conducted using MultiLab 2000 (Thermo Scientific, USA). TFC membranes were surveyed and continuously recorded with the range from 0 to 1100 eV with specific spot size of 650 μm.

2.2.3. Water contact angle analysis. The surfaces of pristine substrate and all modified TFC membranes were appraised by contact angle drop shape geometry (DSA100, Germany) using Milli-Q deionized water as the probe liquid at room temperature. To curtail experimental error, for all TFC membrane samples, the contact angle was randomly measured at more than 5 different locations and the average value was reported.

2.3 Performance of TFC membranes for water vapor permeation

A schematic of the mixed gas vapor permeation set-up for the measurement of gas/water vapor mixtures is shown in ESI Fig. S2.† A mixed water vapor/gas permeation measurement apparatus was developed and has been used by our laboratory for the last few years. A detailed discussion regarding experimental set-up is discussed in the ESI† file. The relative humidity (RH) of the mixed gas flowing through the permeation cell was kept at steady state by controlling distilled water flow rate from the HPLC pump and the constant proportion of carrier gas to dilution gas flow rate. A demister was used to remove the water droplets from the mixed gas. To provide the required vacuum (0.2 bar), which minimizes the pressure drop, a vacuum pump was connected to the permeate stream during the permeation tests. A back pressure regulator was connected to the retentate side so as to operate under the pressure of 2.94 bar. An HMT (humidity and temperature transmitter, probe type 344 Vaisala Oyj, Finland) was used to measure the relative humidity of the feed and retentate streams. Permeation test was carried out inside the homoeothermic oven at a constant temperature, 30 °C. The mixed gas was separated through the TFC membrane, and nitrogen gas from permeate and retentate streams was measured by a bubble flow meter (Gilibrator, USA) after removing the contained water vapor through a cold trap bath (CTB-10, JEIO Tech., Korea). Evaluation of water vapor permeance was carried out using specialized calculation, as shown in eqn (1).

Q_N2 (cm³ s⁻¹) from each of the retentate and the permeate streams is measured by a bubble flow meter after the ice cold trap. γ_H2O is the absolute humidity (g m⁻³) and V_m is the volume of 1 mol penetrant at standard temperature and pressure (22.4 L mol⁻¹). M_w,H2O is the molecular weight of water (18 g mol⁻¹) and Q_vapor (cm³(STP) s⁻¹) is finally determined.

For TFC membranes, permeance (P_i/l) is used to evaluate their performances because the thickness of the thin selective layer cannot be accurately measured. Permeance of component i in the mixture gas was determined by eqn (2).

Permeance (P_i/l) is commonly obtained in gas permeation unit (GPU, 1 GPU = 10⁻⁶ cm³(STP) cm⁻² s⁻¹ cm⁻¹ Hg). N_i is the flux (L m⁻² h⁻¹) at permeate stream through the film by dividing Q_i by the effective area of film membrane A (cm²). Δp_i of component i is the partial pressure difference (cm Hg) between the feed and the permeate side. The partial pressures, p_i,feed and p_i,permeate, on the feed and permeate sides are calculated by multiplying total pressure by mole fraction under the same temperature and pressure. The selectivity (α_i/j) of the membrane for gas i is the ratio of their gas permeance and is calculated by eqn (3).

3. Results and discussion

3.1 Intrinsic properties of TFC membranes

3.1.1. ATR-FTIR. Fig. 4 shows the FTIR absorbance spectra of the PES substrate and TFC membranes to find the effect of different concentrations of PDA solution at a constant reaction time and with 0.5 wt% 3,5-DABA and 0.2 wt% TMC as an aqueous and organic phase monomer solution, respectively. The reaction can occur between amine and carboxylic acid of 3,5-DABA with TMC. If amine groups react with TMC, the polyamide peak would be higher. However, if a carboxylic acid group reacts with TMC, an ester may be formed in the selective layer. As a result, as shown in Fig. 4, a distinct additional peak at 1724 cm⁻¹ can be considered as the C=O of the ester because the absorbance was higher than the lower N–H amide peak. The peaks at 1600 cm⁻¹ and 1492 cm⁻¹ are attributed to the overlap of the C=C resonance vibration in the aromatic ring and the peaks at 1365 cm⁻¹ (phenolic O–H bending) and 1170 cm⁻¹ (phenolic C–O stretching) in PDA are attributed to the N–H bending vibration.^25,26 These new absorption peaks confirm the existence of the polydopamine layer on the PES substrate membrane. These two peaks (1600 cm⁻¹ and 1492 cm⁻¹) were assigned to the overlapping peaks from the C–C vibrations of the aromatic ring and N–H stretching in all the TFC membranes. In TFC membranes, IR peaks at 1545 cm⁻¹ and 1668 cm⁻¹ were obtained due to stretch vibrated N–H and C=O of amide, respectively. The band intensity at 1668 cm⁻¹ increased with an increase in the polydopamine sublayers deposited.²⁷ Bending vibration of the N–H amide appeared at 1627 cm⁻¹. Spectra at 772 cm⁻¹ were detected due specifically to the vibration of the C–Cl stretch, which consists of the unreacted and remaining carbonyl chloride in three-functional trimesoyl chloride. The peaks at 927 and 1724 cm⁻¹ corresponds to the O–H bending, C=O stretching in carboxylic acid, and 1302 cm⁻¹ corresponds to the carbon stretch (C–O) connected to the –OH in carboxylic acid. This can prove that the hydrolysis of –COCl occurred during polymerization. The absorbance peak at 1302 cm⁻¹ is considered close to the carbon stretch (C–O) connected to the –OH in carboxylic acid.

Fig. 4 ATR-FTIR absorbance spectra of PES substrate and thin film composite membranes.

3.1.2. XPS analysis. Chemical components of TFC membranes were analyzed by XPS for different concentrations of PDA coating after formation of the TFC layer by an interfacial polymerization method. As the mole ratio of an atomic element can generally contrast with theoretical values by repeating units,²⁸ we expected cross-linked structures containing both amide and ester functional groups. To examine the spectra for binding energy of TFC membranes, deconvolutions of C1s and O1s by curve fitting are shown in Fig. 5. This analysis by data fitting was based on the data supplied in a previous reference.²⁹

Fig. 5 XPS analysis spectra of TFC thin film composite membranes.

Fig. 5 depicts the distinct high resolution C1s peaks of TFC membranes. All membranes had major peaks at 290 eV. The major peaks are ascribed to C–C/C–H and O=C–O and the major peak of 285 eV is C–H. The minor peak of 288 eV is obtained for C=O. Fig. 5 shows the results of O1s deconvolution by curve fitting for all TFC membranes and it illustrates the change of peaks at some specific binding energy for different wt% PDA coated membranes. Membrane MS1 includes 533.5 eV, MS2 membrane 532.2 eV, MS3 membrane 532.7 eV, and MS4 membrane 533.7 eV obtained due to –(C=O*)–O–(C=O*)–, which is a part of the cross-linked ester groups connected to aromatic rings.

3.1.3. Water contact angle analysis. Water contact angles of the modified surface of TFC membranes were measured and depicted in Fig. 6. To investigate the wetting properties of the modified polyethersulfone membranes, water contact angle was measured by the sessile drop technique. In the case of porous membranes, contact angle is not an efficient way to estimate hydrophilicity because the surface porosity of the substrate readily affects the contact angle by capillary action.³⁰ However, contact angle is a clear indicator of the surface interfacial energy between the thin film and the water droplet. The lesser the contact angle, the higher is the interfacial energy, which results in higher wetting and as a result, high hydrophilicity. A gradual decrease in the contact angles was observed with increasing concentration of PDA. As shown in Fig. 6, the use of 2.0 wt% PDA solution results in the lowest contact angle, 51.2°, for the MS4 membrane sample. The main reasons for this changing contact angle are the thin film selective membrane containing carboxylic acid and hydroxyl groups, and amine functional groups of polydopamine are thought to contribute to the hydrophilicity of coated surfaces.^31–33 The selective polyamide layer of all membranes used in this study were synthesized via the interfacial polymerization of 3,5-diaminobenzoic acid with trimesoyl chloride. The polymerization leaves amine and carboxylic acid groups that are likely responsible for the hydrophilic nature of the tenant membranes.³⁴

Fig. 6 Water contact angle of PDA@PES substrates varied by PDA concentration.

3.2 Effect of PDA concentration and cross-linked structures on water vapor permeation and its selectivity by TFC membranes

The performances of different PDA coated hollow fiber membrane substrates along with 3,5-DABA and TMC monomers at constant concentrations and reaction time are displayed in Fig. 7 and 8. TFC membranes were fabricated by a sequence of modification processes, including PDA coating and formation of a polyamide selective layer on the surface of PDA coated hollow fiber membranes. Water vapor permeance and water vapor/N₂ selectivity were determined using eqn (1)–(3) and discussed earlier. On the one hand, as the experiments were performed under the same experimental conditions, the effect of PDA concentration on TFC membrane performance can be visually observed. As PDA concentration increased, selectivity improved. On the other hand, as one may expect from a diffusion point of view, the membrane also favoured permeation of water vapor over N₂ because of the larger kinetic diameter. The fluxes of water vapor increases with an increase in the PDA sublayers concentration on the HF membrane, and water flux becomes significantly higher.²⁷ Fig. 7 shows the water vapor permeation flux of the thin film composite membranes embracing polyamide and polydopamine coating layers for the separation of water vapor from a gas mixture at 30 °C. Interestingly, adding a polydopamine coating layer on the membrane surface substantially improved membrane selectivity. For instance, when the TFC membrane was deposited on polydopamine, while changing the concentration of PDA, the water content in the permeate side increased; as a result flux increased up to 0.26 kg m⁻² h⁻¹ at 2 wt% PDA solution concentration. It is now clear that changing the concentration of PDA during coating layer preparations yields good selectivity in water vapor/gas separation experiments.

Fig. 7 Results of permeance vs. flux (kg m⁻² h⁻¹) of water vapor for TFC membranes at different PDA coating conditions at constant concentration of monomers (3,5-DABA 0.5 wt%, and TMC 0.2 wt%).

Fig. 8 Results of permeance vs. selectivity of water vapor for TFC membranes at different PDA coating conditions at a constant concentration of monomers (3,5-DABA 0.5 wt% and TMC 0.2 wt%).

As shown in Fig. 7, for the selective layer at higher PDA concentration, it was concluded that membrane performance was readily more dependent on hydrophilicity and inclination of packed structures and morphologies than on thickness of the selective layer because changes in flux as well as permeance of water vapor were equal as a function of 3,5-DABA. As a result, there was no decrease of flux because the PDA coating time for all membrane was only 30 min. Water flux definitely decreases with increase in the coating time period.³⁵ Apart from results in Fig. 8, TFC membranes prepared with different concentrations of PDA coating showed that selectivity had an almost escalating trend at a constant reaction time. It was observed that selectivity gradually improved as PDA coating concentration increased up to a certain level, after that definitely permeance will decrease and only selectivity will increase because the thickness of the PDA layer is responsible for making a dense layer.

The data shown in Fig. 8 demonstrate that the increasing polydopamine concentration for coating the membrane surface led to an increased selectivity of water vapor. The increase in membrane permselectivity was achieved by deposition of the PDA coating layer.²⁷ Prepared TFC membranes are coated on the outer side and contain more hydrophilic groups, which is favorable for water vapor passing through the selective layer of the membrane. The water vapor/N₂ selectivity of TFC membranes was calculated from the water vapor and nitrogen permeabilities, as shown in Fig. 8. In Fig. 8, the mixed vapor/gas selectivity and permeability increased because nitrogen permeability generally decreases with increasing water vapor activity. Piroux et al. found the same,³⁶ although a more prominent effect was observed for the permeation of hydrogen in the presence of water vapor through sulfonated copolyimide membranes. In all the experiments, the temperature was 30 °C. An increase in temperature results in an increase in nitrogen permeability, which is due to the effect of temperature on diffusivity, reducing selectivity.³⁷ Another reason for increasing water vapor permeability is increasing water vapor activity, most likely due to an increased water sorption at higher activities. For all TFC membranes, selectivity increases with increasing PDA concentration due to the increase in water vapor permeability and a decrease in nitrogen permeability. The reduction in selectivity with increase in temperature due to the increased nitrogen permeability at a higher temperature is a well-known mechanism.³⁷ As shown in Fig. 8, for selectivity, an increase of PDA concentration has the ability to improve both the permeance and selectivity.

Our results are dependable with the true cross-linking structure during the reaction between 3,5-DABA and TMC and confirm that the transport properties of thin film composite membranes can be tuned while maintaining defect-free, mechanically strong films necessary for real practical implementation. As shown in Fig. 8, there are two possible causes of increasing permeability and selectivity, one is that disorderly packing of the rigid bulky polymer chain of PDA apparently increases accessible free volume in the polymer matrix without introducing cavities large enough to promote weakly selective or nonselective free-phase flow mechanisms (such as Knudsen or Poiseuille transport). This increase in free volume enhances molecular diffusion coefficients and decreases the size-sifter nature of crosslinked polyamide/polyester, factors that increase both the permeability and selectivity.³⁸ Another possibility for increasing permeance and selectivity is increasing polymer backbone rigidity, resulting in increased selectivity but lower diffusivity and therefore, permeability. This means backbone rigidity increases should be coupled with increases in interchain separation to achieve both higher permeability and higher selectivity.³⁹ These effortless contemplations suggest that simultaneous polymer chain rigidity and interchain separation increases can be used to steadily improve separation performance until the interchain separation becomes hefty enough that the polymer segmental motion no longer affects the penetrant diffusion. As designated above, unless considerable enhancement in selectivity could be achieved, this limit would represent the asymptotic end point in the performance of polymeric membranes.³⁹ Permeance will increase up to certain level, after which it definitely will decrease because of the thickness of the PDA layer being responsible for making a denser layer.

Comparative data for the polymeric membranes in this study and in the literature are listed in Table 3. While taking into consideration the data shown in Table 3, preceding data, and present work results, it is clearly found that the TFC membrane prepared by 3,5-DABA/TMC after PDA coating shows better selectivity. One main thing is when we compare the membrane prepared by the physical and chemical coating method, the stability of the chemical coated membrane is always better, along with membrane performance. Many of the comparative permeability and selectivity values depicted in Table 3 are obtained from pure gas permeabilities by calculating the ratio of the permeabilities for each species.

Table 3 Comparison of water vapor permeability and selectivity for various polymeric membranes

Polymer	Temperature (°C)	Water vapor permeability (GPU)	Selectivity	Reference
a Polysulfone (PSf)/polydopamine (PDA)/thin film composite (TFC).b Polyethersulfone (PES)/cellulose acetate (CA)/polyethylene glycol (PEG).c Polyacrylonitrile (PAN).d Polysulfone (PSf)/thin film composite (TFC).e Polyetherimide (PEI)/poly(amide-b-ethylene oxide) (PEBAX) 1657.
PSf/PDA/TFC^a	30	1029	38 [H2O/N2]	35
PES/CA + PEG2000^b (5 wt%)	30	444	176 [H2O/N2]	40
PAN^c nanofiber-1	50	13 498	27 [H2O/O2]	41
PAN^c nanofiber-2	50	3143	387 [H2O/O2]	41
PSf/TFC^d	30	2160	34 [H2O/N2]	42
PEI/PEBAX1657^e	21	260	274 [H2O/N2]	43
PES/MS1	30	1932	75 [H2O/N2]	This work
PES/MS2	30	2147	105 [H2O/N2]	This work
PES/MS3	30	2660	142 [H2O/N2]	This work
PES/MS4	30	3185	195 [H2O/N2]	This work

---

As the PDA coating layer increased using 0.25 wt% to 2.0 wt% PDA solutions, permeance increased from 1932 GPU to 3185 GPU and selectivity increased from 75 to 195 because the polymerized coating layer by strongly-adhesive PDA can seem to form another layer, which may behave as resistance.³⁵ The trend of water vapor permeance until 60 seconds reaction time was attributed to the crosslinking, which brings improvement in hydrophilicity on the active layer of the TFC membrane as compared to hydrophobic pristine substrates. The increase of selectivity, in spite of the increased permeance, was attributed to the synthesized polyamide, which inhibits nitrogen from permeating through the membrane. In Fig. 8, showing the high permeance of water vapor, it is considered that an increase of permeance while using 2.0 wt% PDA coating solution is ascribed to the IP time during the reaction between –COOH in DABA and hydrolyzed –COCl in TMC, resulting in formation of an ester, which is also somewhat hydrophilic and as a result increases both permeance and selectivity. The high permeance obtained through the membranes is caused by the existence of the unconsumed carboxylic acid. As the reasons are explained in the contact angle section, it is suggested that the excessive PDA layer is favorable to hydrolyze COCl to COOH and leads to condensation reaction itself for esters, which simultaneously increases selectivity. The possible reasons for increasing the selectivity are well explained in the above discussion.

4. Conclusions

TFC membranes used for permeation of water vapor through water vapor/N₂ mixed gas were prepared by interfacial polymerization after PDA coating as a hydrophilic material. To examine the effects of modification on a PES support layer, first we functionalized the coating layer of PDA and prepared a TFC thin layer on the PDA coating layer using two types of monomers with constant concentration and reaction time. As acquired through the characterization, performances, such as vapor flux, permeance and selectivity of the TFC membrane, were evaluated as a function of conditions for interfacial polymerization. The effects of PDA concentration at constant 3,5-DABA and same reaction time with TMC on membrane performances were mainly evaluated in this study. The possibility that the TFC membrane has amide as well as ester groups due to carboxylic acid by interfacial polymerization was first suggested based on ATR-FTIR and XPS results, as shown in Fig. 4 and 5, respectively. As a result, the highest value of water vapor permeance was 3185 GPU and the highest selectivity in water vapor/N₂ mixed gas was 195 on the 2.0 wt% PDA coated membrane.

Acknowledgements

This study was conducted under the framework of Research and Development Program of the Korea Institute of Energy Research (KIER) (B5-2437).

References

1. GEO-4, United Nations Environmental Programme, http://www.unep.org/geo/geo4/media/, 2007, p. 148.
2. D. Wang, Transport membrane condenser for water and energy recovery from power plant flue gas-final technical report, U.S. Department of Energy, Illinois, 2012.
3. B. C. Shin, H. Kwak and K. M. Lee, Korean Chem. Eng. Res., 2012, 50, 646–653.
4. S. H. Hong, S. H. Yoon and J. H. Kim, A Study on the Computer Aided Design of Shell & Tube Type Heat Exchanger, Ocean Industry Research Institute, Korea Maritime, 1996, vol. 5, pp. 95–109.
5. G. Q. Chen, S. Kanehashi, C. M. Doherty, A. J. Hill and S. E. Kentish, J. Membr. Sci., 2015, 487, 249–255.
6. M. Strand, J. Pagels, A. Szpila, A. Gudmundsson, E. Swiethlicki, M. Bohgard and M. Sanati, Energy Fuels, 2002, 16, 1499–1506.
7. X. H. Liu, Y. Zhang, K. Y. Qum and Y. Jiang, Energy Convers. Manage., 2006, 47, 2682–2692.
8. L. Jia, X. F. Peng, J. D. Sun and T. B. Chen, Heat Tran. Asian Res., 2001, 30, 571–580.
9. A. Ito, J. Membr. Sci., 2000, 175, 35–42.
10. Y. H. Zurigat, M. K. Abu-Arabi and S. A. Abdul-Wahab, Energy Convers. Manage., 2004, 45, 141–155.
11. K. Kneifel, S. Nowak, W. Albrecht, R. Hilke, R. Just and K.-V. Peinemann, J. Membr. Sci., 2006, 276, 241–251.
12. L. Z. Zhang, Sep. Sci. Technol., 2006, 41, 1565–1582.
13. L. D. Jia, X. Xu, H. Zhang and J. Xu, J. Appl. Polym. Sci., 1996, 60, 1231–1237.
14. Z. G. Wang, T. L. Chen and J. P. Xu, Macromolecules, 2001, 34, 9015–9022.
15. M. A. M. Beerlage and A. G. L. Zeijsink, Patent WO 0,056,426, 2000.
16. J. Neel, Introduction to pervaporation in: pervaporation membrane, sep. process, Elsevier, Amsterdam, 1991.
17. B. Bolto, M. Hoang and Z. Xie, Water Res., 2012, 46, 259–266.
18. S. R. Reijerkerk, R. Jordana, K. Nijmeijer and M. Wessling, Int. J. Greenhouse Gas Control, 2011, 5, 26–36.
19. H. Lin, S. M. Thompson, A. Serbanescu-Martin, J. G. Wijmans, K. D. Amo, K. A. Lokhandwala, B. T. Low and T. C. Merkel, J. Membr. Sci., 2013, 432, 106–114.
20. J. W. Rhim, H. Y. Hwang, S. Y. Ha and S. Y. Nam, J. Membr. Sci., 2004, 14, 1–17.
21. J. Potreck, K. Nijmeijer, T. Kosinski and M. Wessling, J. Membr. Sci., 2009, 338, 11–16.
22. H. S. Lee, S. M. Dellatore, W. M. Miller and P. B. Messersmith, Science, 2007, 318, 426.
23. K. H. Kim, P. G. Ingole, J. H. Kim and H. K. Lee, Chem. Eng. J., 2013, 233, 242–250.
24. P. G. Ingole, W. Choi, K. H. Kim, C. H. Park, W. K. Choi and H. K. Lee, Chem. Eng. J., 2014, 243, 137–146.
25. B. D. McCloskey, H. B. Park, H. Ju, B. W. Rowe, D. J. Miller and B. D. Freeman, J. Membr. Sci., 2012, 413–414, 82–90.
26. S. Kasemset, A. Lee, D. J. Miller, B. D. Freeman and M. M. Sharma, J. Membr. Sci., 2013, 425–426, 208–216.
27. D. Wu, J. Martin, J. Du, Y. Zhang, D. Lawless and X. Feng, J. Membr. Sci., 2015, 493, 622–635.
28. Z. Y. Xi, Y. Y. Xu, L. P. Zhu, Y. Wang and B. K. Zhu, J. Membr. Sci., 2009, 327, 244–253.
29. G. Beamson and D. Briggs, High Resolution XPS of Organic Polymers, Willey Interscience, 1992.
30. N. Y. Yip, A. Tiraferri, W. A. Phillip, J. D. Schiffman and M. Elimelech, Environ. Sci. Technol., 2010, 44, 3812–3818.
31. H. C. Yang, J. Luo, Y. Lv, P. Shen and Z. K. Xu, J. Membr. Sci., 2015, 483, 42–59.
32. J. H. Jiang, L. P. Zhu, X. L. Li, Y. Y. Xu and B. K. Zhu, J. Membr. Sci., 2010, 364, 194–202.
33. L. P. Zhu, J. Z. Yu, Y. Y. Xu, Z. Y. Xi and B. K. Zhu, Colloids Surf., B, 2009, 69, 152–155.
34. E. M. van Wagner, A. C. Sagle, M. M. Sharma, Y. H. La and B. D. Freeman, J. Membr. Sci., 2011, 367, 273–287.
35. S. H. Yun, P. G. Ingole, K. H. Kim, W. K. Choi, J. H. Kim and H. K. Lee, Chem. Eng. J., 2014, 258, 348–356.
36. F. Piroux, E. Espuche and R. Mercier, J. Membr. Sci., 2004, 232, 115–122.
37. H. Sijbesma, K. Nymeijer, R. V. Marwijk, R. Heijboer, J. Potreck and M. Wessling, J. Membr. Sci., 2008, 313, 263–276.
38. T. C. Merkel, B. D. Freeman, R. J. Spontak, Z. He, I. Pinnau, P. Meakin and A. J. Hill, Science, 2002, 296, 519–522.
39. B. D. Freeman, Macromolecules, 1999, 32, 375–380.
40. K. H. Kim, P. G. Ingole, S. H. Yun, W. K. Choi and H. K. Lee, J. Chem. Technol. Biotechnol., 2015, 90, 1117–1123.
41. R. Huizing, W. Mérida and F. Ko, J. Membr. Sci., 2014, 461, 146–160.
42. S. H. Yun, P. G. Ingole, K. H. Kim, W. K. Choi and H. K. Lee, J. Mater. Chem. A, 2015, 3, 7888–7899.
43. H. Lin, S. M. Thomson, A. Serbanescu-Martin, J. G. Wijmans, K. D. Amo, K. A. Lokhandwala and T. C. Merkel, J. Membr. Sci., 2012, 413–414, 70–81.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra15199f

---