Hyper-Cross-Linked Additives that Impede Aging and Enhance

Research Article www.acsami.org

Hyper-Cross-Linked Additives that Impede Aging and Enhance Permeability in Thin Polyacetylene Films for Organic Solvent Nanofiltration Xi Quan Cheng,†,‡ Kristina Konstas,† Cara M. Doherty,† Colin D. Wood,§ Xavier Mulet,† Zongli Xie,† Derrick Ng,† Matthew R. Hill,*,†,∥ Lu Shao,*,‡ and Cher Hon Lau*,†,⊥ †

CSIRO, Private Bag 10, Clayton South, Victoria 3169, Australia MIIT Key Laboratory of Critical Materials Technology for New Energy Converson and Storage, State Key Laboratory of Urban Water Resource and Environment, School of Chemistry and Chemical Engineering, Harbin Institute of Technology, Harbin 150001, China § CSIRO, Australian Resources Research Centre, Kensington, Western Australia 6155, Australia ∥ Department of Chemical Engineering, Monash University, Clayton Victoria 3800, Australia ⊥ Department of Chemical Engineering, University of Edinburgh, Edinburgh EH9 3JL, United Kingdom

Downloaded via TULANE UNIV on January 20, 2019 at 14:57:01 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

‡

ABSTRACT: Membrane materials with high permeability to solvents while rejecting dissolved contaminants are crucial to lowering the energy costs associated with liquid separations. However, the current lack of stable high-permeability materials require innovative engineering solutions to yield high-performance, thin membranes using stable polymers with low permeabilities. Poly[1-(trimethylsilyl)-1-propyne] (PTMSP) is one of the most permeable polymers but is extremely susceptible to physical aging. Despite recent developments in anti-aging polymer membranes, this research breakthrough has yet to be demonstrated on thin PTMSP films supported on porous polymer substrates, a crucial step toward commercializing anti-aging membranes for industrial applications. Here we report the development of scalable, thin film nanocomposite membranes supported on polymer substrates that are resistant to physical aging while having high permeabilities to alcohols. The selective layer is made up of PTMSP and nanoporous polymeric additives. The nanoporous additives provide additional passageways to solvents, enhancing the high permeability of the PTMSP materials further. Through intercalation of polyacetylene chains into the sub-nm pores of organic additives, physical aging in the consequent was significantly hindered in continuous long-term operation. Remarkably we also demonstrate that the additives enhance both membrane permeability and rejection of dissolved contaminants across the membranes, as ethanol permeability at 5.5 × 10−6 L m m−2 h−1 bar−1 with 93% Rose Bengal (1017.6 g mol−1) rejection, drastically outperforming commercial and stateof-the-art membranes. These membranes can replace energy-intensive separation processes such as distillation, lowering operation costs in well-established pharmaceutical production processes. KEYWORDS: organic solvent nanofiltration, physical aging, porous aromatic frameworks (PAF-1), hyper-cross-linked polymer (HCP), solvent treatment

■

INTRODUCTION

organic solvent nanofiltration (OSN) membranes are preferred.1,4 Such membranes are characterized by their ability to prevent certain molecular weight compounds from permeating and their solvent-resistance. OSN membranes are typically fabricated using glassy polymers with low intrinsic permeabilities5 such as polyamide (PA), polyimide (PI), and polyelectrolytes (PEs). These polymers are chosen for their high rejection performance and excellent performance stability (permeance or rejection).2,6−8 Clearly, the choice of glassy polymers for OSN membranes is

Alcohols are a valuable commodity material and an important chemical building block and are widely used for purifying valuable chemical and pharmaceutical products in the industry to minimize product waste. Depending on the end application, alcohol purification can also be designed to maximize product recovery.1,2 Batch solvent purification techniques, such as distillation and molecular sieving, are fundamental processes for the production of alcohols despite their high energy and spatial requirements. A green alternative for these energy-intensive purification techniques is membrane separation techniques (including pervaporation, membrane distillation) where energy consumption and investment costs are significantly lower.3 In applications where alcohols are isolated from solid molecules, © 2017 American Chemical Society

Received: February 16, 2017 Accepted: April 4, 2017 Published: April 4, 2017 14401

DOI: 10.1021/acsami.7b02295 ACS Appl. Mater. Interfaces 2017, 9, 14401−14408

Research Article

ACS Applied Materials & Interfaces

Figure 1. (A) Chemical structures of PTMSP, p-DCX, and PAF-1. (B) As-cast PTMSP membranes cast on porous polycarbonate substrates are selective toward alcohols over molecular dyes. After continuous operation for 500 h, PTMSP membranes are no longer selective. (C) The incorporation of porous additives such as PAF-1 or p-DCX into PTMSP enhanced the permeance of alcohols and remained permeable and selective toward alcohols. (D) The average thicknesses of membranes studied here is ∼1 μm; the standard deviation of membrane thickness is ±10% (measured using SEM).

which is about two magnitudes higher than that of commercial polyimide membranes, Starmem240 with similar rejection.16 Interestingly, the permeance of 35 nm-thick of PIM-1 is much lower than that of the 140 nm of PIM-1, implying that the loss of FFV content in thinner membranes is more significant.16 Crucially, physical aging in super glassy polymers must be overcome to exploit their high intrinsic permeances to develop ultrapermeable OSN membranes in a sustainable, green manner. A simple approach that simultaneously overcomes both physical aging and low alcohol permeance is to deploy mixed matrix membranes that do not age. Super glassy polymers and nanoporous additives such as porous aromatic frameworks, PAF-1,18 or a hyper-cross-linked polymer such as polydichloroxylene (p-DCX)19 can provide longevity while enhancing gas transport in PTMSP. The key to anti-aging PTMSP membranes is the immobilization of polymer chains through their compatibility with similar functional groups on specific porous frameworks (non-bonding interactions). The anti-aging capabilities of these mixed matrix membranes were demonstrated using 100 μm thick films. The physical aging rates of polymer films with various thicknesses are different as thin films age faster than thick films.20 More importantly, these anti-aging mechanisms were observed in batch experiments where low pressure gases were only exposed to these thick films periodically. Although these experiments demonstrated the world’s first anti-aging membranes, the performance of these membranes in continuous operating conditions remain unknown. Alternative to using these anti-aging membranes for batch gas separation processes, we deployed 1 μm thick PTMSP composite membranes for continuous organic solvent nanofiltration (OSN). It is important to note that the separation mechanism of gas separation and nanofiltration are usually different so it is not intuitive that this approach would work. The gas separation mechanism in our anti-aging, dense film membranes can be described with the solution-diffusion model,21 while the separation mechanism for most OSN membranes is based on the pressure-driven model.1,6 Because of their intrinsic porosity, PAF-1 and p-DCX will further

limited due to the trade-off between membrane permeance, rejection performance, and performance stability. This hampers the widespread application of OSN membranes. To enhance the alcohols permeance of OSN membranes without compromising rejections, the selective membrane layer should be as thin as possible (reducing thickness) or created with additional solvent passageways (increasing porosity). However, performance instability is omnipresent in thin membranes as alcohol permeances decline over time. This is attributed to physical aging where pore sizes are reduced and pore concentration is reduced (loss of FFV).9−11 For example, the alcohol permeances of ultrapermeable 95 nm thin PA OSN membranes were extremely unstable when supported on polymeric substrates.12 The ultrathin polyamide selective layers coated on polyimide supported membranes exhibit about 50% permeance decline after 4 h continuous operation under 10 bar.12 Physical aging in this ultrathin PA membrane was only overcome by replacing the polymeric support with inorganic supports. The role of substrates in overcoming physical aging in ultrathin films remain unknown. Moreover, the etching of Cd(OH)2 nanostrands sacrificial layer to create additional porosity for ultrapermeable membranes generates heavy metal ions (Cd2+) which is carcinogenic and harmful to the environment. A greener alternative to yield ultrapermeable OSN membranes is to use super glassy polymers such as PIMs and PTMSP.13−16 The key advantage of these polymers is the exceptionally high FFV content, i.e., additional porosity for molecular transport.13−16 The high permeability of these membrane materials precludes the need to develop ultrathin membranes that are prone to defects using harmful synthesis procedures. Unfortunately, membranes fabricated with high FFV materials suffer from serious physical aging. During physical aging, polymer chains that are propped further apart by kinks or bulky functional groups converge toward a thermodynamic equilibrium, collapsing the free space, i.e., FFV between polymer chains. This is more prevalent in ultrathin-membranes.16,17 For example, 140 nm thin PIM-1 membranes demonstrated heptane permeances as high as 18 L m−2 h−1 bar−1(permeability as 2.8 × 10−6 L m m−2 h−1 bar−1), 14402


Research Article


Figure 2. Chemical structures of molecular dyes used in this work.

Figure 3. Fabrication of thin film PTMSP, PTMSP/PAF-1, and PTMSP/p-DCX membranes. without purification. Chloroform, dichloromethane (DCM), and hydrochloric acid (HCl) were used as received. 1,4, Dichloroxylene (DCX) 98%, iron(III) chloride, reagent grade, anhydrous 97%, tetraphenylmethane, bis(1,5-cyclooctadiene) nickel, and 2,2′-bipyridyl were purchased from Sigma-Aldrich and used without further purification. 1,2 Dichloroethane (DCE) was supplied by ChemSupply. The polycarbonate porous substrate was commercial microfiltration membranes (Nuclepore) with mean effective pore size of 0.2 μm and ethanol permeances above 500 L m−2 h−1 bar−1 provided by Whatman, GE Company. Synthetic Procedures. PAF-1 Synthesis. PAF-1 was synthesized according to Zhu and co-workers22 to yield an off-white powder with a BET surface area of 3760 m2/g). 1,5-Cyclooctadiene (dried over CaH2) was added into a solution of bis(1,5-cyclooctadiene) nickel and 2,2′-bipyridyl in dehydrated DMF and heated at 80 °C to form a purple solution. Tetrakis(4-bromophenyl)methane was added to the mixture and stirred overnight at 80 °C. The mixture was allowed to cool to room temperature and concentrated HCl was added. The solids were collected and washed with chloroform, THF, and deionized water. The particle size was typically in the range of 100− 200 nm characterized by a nanoparticle size analyzer.19 p-DCX Synthesis. To a solution of DCX monomer (0.171 mol, 30 g) in anhydrous DCE (388 mL), a DCE solution (388 mL) of FeCl3 (0.173 mol, 28 g) was added. The resulting mixture was stirred in an open vessel at room temperature. The precipitated p-DCX was washed once with water, three times with methanol (until the filtrate was clear), and with diethyl ether followed by drying for 24 h at 60 °C. The

enhance the solvent permeabilities and possibly, dye rejections and minimize physical aging in 1 μm thin PTMSP membranes supported on porous polymeric (polycarbonate) substrates for OSN. Our approach removed the need to sacrifice initial membrane performances for stability or complex engineering means to achieve fast alcohol permeance (Figure 1). In order to mimic industrial pharmaceutical recovery from alcohols, a continuous long-term membrane operation in OSN conditions (5 bar, 25 °C, stirring speed of 700 rpm) was carried out and revealed that these porous additives drastically enhanced alcohol permeances of PTMSP membranes, increased stability, improved dye rejections, and provided resistance against compaction. Overall the membrane separation performances are “locked” in their initial state which justifies revisiting super glassy polymers for high speed alcohol purification at low operating pressures as our anti-aging, thin PTMSP composite membranes demonstrate higher separation efficiency than most state-of-the-art OSN membranes.

■

EXPERIMENTAL SECTION

Chemicals and Materials. Analytical grade methanol, ethanol, isopropanol, Rose Bengal (RB), Crystal Violet (CV), and Thiazole Yellow (TY) (Sigma-Aldrich) (Figure 2) were purchased and used without further purification. PTMSP was purchased from Gelest Inc. (SSP070-10GM, Lot 4I-23599, Mw 210 kDa, 95% purity) and used 14403


Research Article

ACS Applied Materials & Interfaces BET surface area of the p-DCX nanoparticles is 1330 m2/g; the diameter of p-DCX nanoparticles is between 20 and 50 nm.19 Membrane Preparation. The polymer solution was prepared by dissolving 2 wt % of PTMSP (Gelest Inc.) in cyclohexane. The 10 wt % (with respect to PTMSP concentration) of porous additives, PAF-1 or p-DCX, were added to the polymer solution and stirred for 24 h. This mixture was then poured onto a glass plate. Doctor blades with thicknesses 30, 50, 80, 100, 150 μm were used as cast membranes. Upon solvent evaporation, the glass plate with a thin layer of membrane was lowered into a water bath. This enabled the flotation of the thin polymer membrane film onto the surface of the water. A polycarbonate porous substrate was lowered into the water bath and came into contact with the PTMSP-based membrane. Water was then drained out, and a free-standing PTMSP-based membrane supported on a polycarbonate substrate was obtained (Figure 3). Because of similarities in chemical structure, PAF-1 and p-DCX are well-dispersed in the PTMSP matrix.19

■

Figure 4. Pore size distribution of PAF-1 (blue) and p-DCX (red) nanoparticles obtained from nitrogen adsorption isotherms performed at 77 K. and the range of the TAC extended to 200 ns. The film samples were stacked to a thickness of 2 mm, and powdered samples were packed (>1.5 mm depth) on either side of a 30 μCi 22NaCl source sealed in a 2.54 μm thick Mylar envelope (source correction 1.605 ns and 2.969%) and measured at ∼5 × 10 −7 Torr. At least five spectra of 4.5 million integrated counts were collected with each spectrum taking about 4.6 h to collect. Data analysis was performed using LT9. The spectra were best fitted with five components with the shortest lifetime fixed to 125 ps, characteristic of p-Ps annihilation. For the long lifetimes obtained, the Tao-Eldrup model23,24 traditionally used for calculating mean pore sizes from mean o-Ps lifetimes is not valid; therefore, the mean free path (nm) of the pores was calculated using the Rectangular Tao Eldrup (RTE) model.25 The FFV content of freestanding 1 μm thin membranes are shown in Table 1.

METHODS

Organic Solvent Nanofiltration. The permeances or fluxes of OSN membranes were measured using a self-made, stainless steel dead-end pressure cell with a membrane area of 21.2 cm2. The feed solution was pressurized with nitrogen to 5 bar at room temperature. During filtration, the feed solution was stirred at 11.66 Hz (700 rpm) to avoid concentration polarization. Permeate samples were collected in cooled flasks as a function of time, weighed, and analyzed. The solvent flux and solvent permeance were calculated using the following equations:

F = V /A × t

(1)

permeance = F /ΔP

(2)

permeability = permeance/n

Table 1. FFV Content of Free-Standing Dry and Wet (EtOH Soaked) PTMSP, PTMSP/PAF-1, and PTMSP/p-DCX Membranes

(3) −2

−1

where F represents the solvent flux (L m h ), V (L) is the volume of the solvent (or solution) passing through the membrane, A is the effective membrane area (m2), t is the operation time (h); and ΔP is trans-membrane pressure (bar), n presents the thickness of the membranes (m). Solvent permeances were tested in the order of MeOH > EtOH > i-PrOH. The solute rejections of NF membranes were calculated using eq 4

⎛ Cp ⎞ R = ⎜1 − ⎟ × 100% Cf ⎠ ⎝

sample name

FFV in dry-state (%)

FFV wet-state (%)

PTMSP PTMSP/PAF-1 PTMSP/p-DCX

20 21 30

11 14 12

Contact Angle Measurements. A contact angle measuring system (G10 Kruss, Germany) was used to measure the static water contact angle of membranes. A deionized water droplet was placed on a dry flat membrane surface and the contact angle was obtained. The reported contact angle value was calculated by averaging over more than five contact angle values at different sites. The contact angles of PTMSP and PTMSP/PAF-1 membranes are summarized in Table 2.

(4)

where Cp and Cf are the solute concentrations in the permeate and the feed solution, respectively. Dye concentrations in IPA were measured with a UV−vis CINTRA20-GBC apparatus (λmax of RB = 548 nm). Each data point is an average of three repetitions of each test, with ±5% standard deviation. Scanning Electron Microscopy (SEM). All membrane films were cryo-fractured, to achieve a clean break, and then mounted on crosssection SEM sample stubs. The average thickness for the selective layer is ∼1 μm. Nitrogen Adsorption Isotherms. Gas adsorption isotherms between the range of 0−700 mmHg were measured by a volumetric approach using a Micrometrics ASAP 2420 instrument. All the samples were transferred to pre-dried analysis tubes, sealed with Transeal stoppers, evacuated, and activated at 120 °C under a 10−6 dynamic vacuum for 24 h. Ultrahigh purity N2 gases were used for these experiments. N2 adsorption measurements were conducted at 77 K. The pore size distributions of PAF-1 and p-DCX polymers are shown in Figure 4. Positron Annihilation Lifetime Spectroscopy (PALS). Bulk PALS experiments that were used to characterize the changing of pores and the FFV in the material bulk were performed at CSIRO using an EG&G Ortec fast−fast coincidence system with fast plastic scintillators and a resolution function of 260 ps fwhm (60Co source with the energy windows set to 22Na events). Because of the long lifetimes and the low counting rate, the coincidence unit was removed

Table 2. Contact Angles of PTMSP and PTMSP/PAF-1 Membranes Studied Here in This Work sample name

water contact angle (deg)

PTMSP PTMSP/PAF-1 PTMSP/p-DCX

117 127 122

■

RESULTS AND DISCUSSION Ethanol (EtOH) is commonly used to isolate water-soluble pharmaceuticals1 and is an ideal solvent choice for OSN experiments with different dyes and physical aging tests.13 Since the osmotic pressure is low for purification or ethanol recovery, 5.0 bar of operation pressure is chosen here to minimize energy consumption, precluding the need of a high driving force to achieve high separation efficiencies of our ultrapermeable membranes. The incorporation of PAF-1 and p-DCX into PTMSP drastically enhanced ethanol permeation by 42% and 90%, respectively (Figure 5A), while improving dye rejection. 14404


Research Article


5.9 and 13.5 Å in p-DCX nanoparticles, while there are three overlapping pore size distributions in PAF-1 in the region between 11.8 and 17.2 Å (Figure 4). As the side-chain of the PTMSP can thread the larger pores, larger pores of the PAF-1 and p-DCX were partly blocked, leaving the smaller pores intact.18,19 Alcohol molecules can pass through these unblocked 5.9 Å pores27 in p-DCX, hence contributing to the 90% enhancement in EtOH permeance. The smaller pores of pDCX nanoparticles also accounted for the 15% improvement in rejecting Safranine O, the smallest dye studied here (Mw 350.13 g mol−1). PAF-1 and p-DCX nanoparticles slightly improved the 90% rejection rate of PTMSP membranes for both the larger Rose Bengal (Mw 1017.64 g mol−1) and Thiazol Yellow (Mw 695.74 g mol−1) dyes. Unlike other mixed matrix membranes, the incorporation of porous additives did not reduce dye rejection rates, indicating an absence of large nanogaps between the porous additives and polymer chains.28 The dye rejection rates of present PTMSP/additive membranes are comparable to commercial membranes.29 Remarkably, the permeability of the PTMSP/p-DCX nanocomposite reaches 5.5 × 10−6 L m m−2 h−1 bar−1, which is higher than that of PIM-1 based membranes16 and ultrathin PA membranes.12 Physical aging was investigated using OSN membranes operated continuously for 500 h at 5 bar (Figure 5B). The EtOH permeance in PTMSP control membranes was reduced by 45% and stabilized within the first 100 h of aging. Physical aging reduced the pore sizes and concentration in PTMSP, hence impeding ethanol transport in aged PTMSP membranes. PAF-1 and p-DCX inhibited or abated the shrinkage and loss of such pores in anti-aging18 and selective-aging gas separation PTMSP membranes.19 Here we report that this was also valid for alcohol transport across 1 μm thin PTMSP membranes. The enhanced EtOH permeances of PTMSP membranes loaded with PAF-1 and p-DCX were only reduced by 20 and 12%, respectively, and stabilized within 13 h of testing.

Figure 5. (A) Effects of physical aging (after 500 h of continuous operation at 5 bar of solvent pressure and 25 °C) on ethanol transport in PTMSP (black), PTMSP/PAF-1 (blue), and PTMSP/p-DCX (red). Membranes were cast using 0.5 wt % of doping concentration. (B) Dye rejection rates of as-cast (solid) and aged (empty) membranes studied here. The molecular weights of Rose Bengal, Thiazol Yellow, and Safranine O are 1017.64 g mol−1, 695.74 g mol−1, and 350.13 g mol−1, respectively. Each separation performance data point is obtained from an average of three repeated experiments, with ±6% standard deviation.

The increase in EtOH permeance was attributed to the intrinsic pores of PAF-1 and p-DCX nanoparticles that provided additional molecular transportation pathways,18,19 the hydrophobicity of additives,26,27 and increase in pore size for PTMSP membranes. Highly hydrophobic 5 Å pores in PAF-1 nanoparticles optimize alcohol absorption and alcohol diffusion.27 There are two pore size distributions centered at

Figure 6. Methanol (black), ethanol (red), and isopropanol (blue) permeances of as-cast (solid) and aged (empty) 1 μm thin (A) PTMSP, (B) PTMSP/PAF-1, and (C) PTMSP/p-DCX membranes supported on polycarbonate substrates. The Rose Bengal (orange), Thiazole Yellow (green), and Safranine O (purple) dyes rejection rates from ethanol of (D) PTMSP, (E) PTMSP/PAF-1, and (F) PTMSP/p-DCX membranes. Each separation performance data point is obtained from an average of three repeated experiments, with ±5% standard deviation. 14405


Research Article


bimodal pore size distribution centered at 5.5 Å (d3) and 14 Å (d4) in PTMSP (Figure 7A). Physical aging reduced the d3 and d4 pore sizes in PTMSP by ∼0.5 and 2 Å, respectively. The incorporation of PAF-1 and p-DCX nanoparticles did not alter the pore sizes and concentration in as-cast PTMSP, indicating the good compatibility between PTMSP and the porous additives. With PAF-1 and p-DCX, the d4 pores in aged PTMSP films were only reduced by 1 Å, while the d3 pore size and the d3 and d4 pore concentrations remain unchanged. Aged 1 μm thin PTMSP, PTMSP/PAF-1, PTMSP/p-DCX membranes lost 48, 40, and 20% of FFV content (Figure 7B). Higher FFV content, i.e., porosity in aged PTMSP/additive membranes lowered transmembrane resistance and enhanced total flux.31 The relative losses in FFV content were smaller than relative losses in EtOH permeances in aged PTMSP/ additive membranes. The impact of lost FFV content on EtOH transport could be mitigated by alcohol adsorption. This view is reinforced by the fact that PTMSP/PAF-1 membranes with higher FFV content had lower EtOH permeances when compared to PTMSP/p-DCX membranes with lower FFV content. We also observed changes in pore sizes and concentrations in aged dry and wet (soaked in EtOH) PTMSP-based films (Figure 7C). Polymer films were soaked in EtOH for 24 h, dried with tissue paper, prior to characterization. In the presence of EtOH, the pore sizes in and the concentration of d3 and d4 pores decreased. This indicated that d4 pores were filled up with EtOH molecules. The filling of pores with EtOH molecules was also observed across both d3 and d4 pores in both aged PTMSP/PAF-1 and PTMSP/p-DCX membranes. The filling of d3 pores in PTMSP/additive films could be attributed to film hydrophobicity. PAF-1 and p-DCX nanoparticles increased the hydrophobicity of PTMSP films (Table 2) that improved alcohol adsorption.27 This accounted for the highest EtOH adsorption of 6.4 cm3/g in PTMSP/p-DCX membranes, 17% higher than PAF-1 loaded membranes (Figure 7D). Higher EtOH adsorption did not swell PTMSP/additive membranes but filled up the pores present in these membranes. The lack of swelling due to EtOH adsorption in our PTMSP/additive membranes indicated that the rigidification of PTMSP polymer chains by porous nanoparticles, stabilizing membrane permeances during longterm operation. This was further demonstrated through alcohol regeneration and membrane compaction tests. Alcohol regeneration is used to rejuvenate the collapsed FFV content between mobile polymer chains.32 Here we used this technique to reveal the rigidification of PTMSP chains by PAF1 and p-DCX nanoparticles. Membranes were first exposed to 5 bar EtOH for 100 h and regenerated by a 100 h EtOH soak. EtOH permeance in PTMSP, PTMSP/PAF-1 and PTMSP/pDCX membranes was reduced by 45%, 20%, and 12%, respectively, after 100 h continuous operation (Figure 8). After depressurize of 5 h, the alcohol permeance of aged PTMSP membranes recovered 12%. After 100 h continuous soaking in ethanol, the permeances of the aged PTMSP were further recovered by 33%. Without PAF-1 or p-DCX nanoparticles, PTMSP chains possessed more freedom and mobility.18,19 Through alcohol regeneration, alcohol molecules occupied and recovered the free spaces between these mobile polymers chains.32 Mobile PTMSP chains relaxed upon depressurization and regained some free volume content that contributed to the recovery of molecular transportation rates,33 hence, accounting for the significant recovery of EtOH

Subsequently, the permeances of membranes loaded with pDCX and PAF-1 remained stable during 487 h of continuous operation. The anti-aging effect is more pronounced with pDCX nanoparticles. Different from gas separation membranes where only the permeation of large molecules like nitrogen (kinetic diameter 3.64 Å) were affected by physical aging,18,19 we observed that the permeation of even larger molecules like ethanol (kinetic diameter 4.5 Å30) and i-PrOH (kinetic diameter 4.7 Å)30 (Figure 6) were not affected in our PTMSP/additive OSN membranes. Clearly, there is a different anti-aging mechanism for OSN. Positron annihilation lifetime spectroscopy (PALS) was used to track changes in pore size and concentration and FFV content within 1 μm thin free-standing PTMSP-based films in wet (presoaked in EtOH) and dry states (Figure 7). There is a

Figure 7. (A) Bimodal pore size distribution of as-cast vs aged PTMSP (black), PTMSP/PAF-1 (blue), and PTMSP/p-DCX (red) membranes determined using PALS. (B) The relationship between relative EtOH permeances of PTMSP, PTMSP/PAF-1, and PTMSP/p-DCX membranes and FFV content losses due to physical aging. (C) Pore size distribution of as-cast wet (dotted lines) vs dry (solid lines) PTMSP composite membranes studied here. (D) The influence of additives on the EtOH adsorption in PTMSP membranes studied here. 14406


Research Article


separation process. Addition of p-DCX reduced physical aging rates by 12%, while also doubling the permeance rates of EtOH through the fresh membrane. Careful experiments simulating applied settings delivered over 500 h of continuous operation with stabilized performance, with as much as 90% higher permeance than aged controls. Detailed studies of various alcohols demonstrated the potential use of this system across a platform of low-energy liquid purification applications. The permeabilites of our PTMSP/porous additive membranes stabilized to values that were higher than the initial permeances of as-cast PTMSP membranes, outperforming current state-ofthe-art membranes1,12,13,34−38 (Figure 9). Unique threading of the polymer side chains into the additive pores was found to underpin this remarkable performance. Taken together, these findings enable further use of our compaction-free OSN membranes that retain most of their initial permeances as low energy separation alternatives and further empower the highest performing polymers to be included in these membranes by imbuing them with stable solvent permeabilities.

Figure 8. Incorporation of PAF-1 and p-DCX voids the need to regenerate PTMSP membranes to recover initial EtOH permeances: aged membranes (100 h continuous separation), depressurized membranes means the aged membranes was soaked in ethanol under atmosphere for 5.0 h to recover the deformation of polymer chain under pressure, and regenerated membranes means the aged membranes soaked in ethanol for 100 h.

■

permeance in pure PTMSP membranes. As PTMSP chains were immobilized by PAF-1 or p-DCX nanoparticles, the rigidification of PTMSP chains membranes was enhanced, thereby inhibiting physical aging and subduing alcohol regeneration. Interestingly, even without alcohol regeneration, the EtOH permeances of PTMSP/PAF-1 and PTMSP/p-DCX membranes remained stable and significantly higher than regenerated PTMSP.

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. *E-mail: [email protected]. ORCID

■

Lu Shao: 0000-0002-4161-3861 Cher Hon Lau: 0000-0003-1368-1506

CONCLUSIONS In the preceding, the unique interactions between porous additives such as PAF-1, p-DCX, and the super glassy polymer

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS X.Q.C. and C.H.L. have contributed equally to this work. M.R.H., K.K., and C.H.L. acknowledge the Science and Industry Endowment Fund (SIEF). M.R.H. and A.J.H. acknowledge the generous support of the CSIRO Office of the Chief Executive Science team. M.R.H. acknowledges Grant FT 130100345. C.M.D. is funded through an Australian Research Centre DECRA project (Grant DE140101359). This work was supported by National Natural Science Foundation of China (Grants 21676063, U1462103), the Open Research Fund Program of Collaborative Innovation Center of Membrane Separation and Water Treatment of Zhejiang Province, State Key Laboratory of Urban Water Resource and Environment (Harbin Institute Technology) (Grant No. 2017DX05), and HIT Environment and Ecology Innovation Special Funds (Grant HSCJ201619).

Figure 9. Membranes reported in the literature are of different thicknesses.1,12,13,34−38 Hence we compared the EtOH permeabilities and Thiazol Yellow (Mw 695.74 g mol−1) dye rejection rates of our aged, thin film, composite membranes with other membranes. The molecular weight of most dyes reported in these literature ranged between 422 and 626 g mol−1. The dotted line represented Karan, Jiang, and Livingston’s state-of-the-art polyamide membrane with a sub-10 nm ultrathin selective layer.12 In their work, dye (Mw 246.2− 585.5 g mol−1) rejection experiments were only conducted using methanol.

■

REFERENCES

(1) Marchetti, P.; Jimenez Solomon, M. F.; Szekely, G.; Livingston, A. G. Molecular separation with organic solvent nanofiltration: a critical review. Chem. Rev. 2014, 114, 10735−10806. (2) Szekely, G.; Jimenez-Solomon, M. F.; Marchetti, P.; Kim, J. F.; Livingston, A. G. Sustainability assessment of organic solvent nanofiltration: from fabrication to application. Green Chem. 2014, 16, 4440−4473. (3) Schneiderman, S. J.; Gurram, R. N.; Menkhaus, T. J.; Gilcrease, P. C. Comparative technoeconomic analysis of a softwood ethanol process featuring posthydrolysis sugars concentration operations and continuous fermentation with cell recycle. Bioethanol. Prog. 2015, 31, 946−956.

PTMSP have been utilized to enhance membrane permeabities and dye rejections and reduce physical aging of the membranes. This also ensures that the initial tantalizing separation properties of the membranes are stable in continuous 14407


Research Article

ACS Applied Materials & Interfaces (4) Székely, G.; Marchetti, P.; Jimenez-Solomon, M. F.; Livingston, A. G.; Hoek, E. M. V.; Tarabara, V. V. In Encyclopedia of Membrane Science and Technology; John Wiley & Sons, Inc.: Hoboken, NJ, 2013. (5) Campbell, J.; Davies, R. P.; Braddock, D. C.; Livingston, A. G. Improving the permeance of hybrid polymer/metal−organic framework (MOF) membranes for organic solvent nanofiltration (OSN)− development of MOF thin films via interfacial synthesis. J. Mater. Chem. A 2015, 3, 9668−9674. (6) Vandezande, P.; Gevers, L. E. M.; Vankelecom, I. F. J. Solvent resistant nanofiltration: separating on a molecular level. Chem. Soc. Rev. 2008, 37, 365−405. (7) Ahmadiannamini, P.; Li, X. F.; Goyens, W.; Meesschaert, B.; Vanderlinden, W.; De Feyter, S.; Vankelecom, I. F. J. Influence of polyanion type and cationic counter ion on the SRNF performance of polyelectrolyte membranes. J. Membr. Sci. 2012, 403-404, 216−226. (8) Li, X. F.; De Feyter, S.; Chen, D. J.; Aldea, S.; Vandezande, P.; Du Prez, F.; Vankelecom, I. F. J. Solvent-resistant nanofiltration membranes based on multilayered polyelectrolyte complexes. Chem. Mater. 2008, 20, 3876−3883. (9) Goulas, A. K.; Kapasakalidis, P. G.; Sinclair, H. R.; Rastall, R. A.; Grandison, A. S. Purification of oligosaccharides by nanofiltration. J. Membr. Sci. 2002, 209, 321−335. (10) Schmidt, P.; Kose, Y.; Lutze, P. Characterisation of organic solvent nanofiltration membranes in multi-component mixtures: Membrane rejection maps and membrane selectivity maps for conceptual process design. J. Membr. Sci. 2013, 429, 103−120. (11) Rezzadori, K.; Penha, F. M.; Proner, M. C.; Zin, G.; Petrus, J. C. C.; Prádanos, P.; Palacio, L.; Hernández, A.; Luccio, M. D. Evaluation of reverse osmosis and nanofiltration membranes performance in the permeation of organic solvents. J. Membr. Sci. 2015, 492, 478−489. (12) Karan, S.; Jiang, Z.; Livingston, A. G. Sub−10 nm polyamide nanofilms with ultrafast solvent transport for molecular separation. Science 2015, 348, 1347−1351. (13) Volkov, A. V.; Parashchuk, V. V.; Stamatialis, D. F.; Khotimsky, V. S.; Volkov, V. V.; Wessling, M. High permeable PTMSP/PAN composite membranes for solvent nanofiltration. J. Membr. Sci. 2009, 333, 88−93. (14) Volkov, A. V.; Tsarkov, S. E.; Gokzhaev, M. B.; Bondarenko, G. N.; Legkov, S. A.; Kukushkina, Y. A.; Volkov, V. V. Nanofiltration and sorption of organic solvents in poly(1-trimethylsilyl-1-propyne) samples of different microstructures. Pet. Chem. 2012, 52, 598−608. (15) Fritsch, D.; Merten, P.; Heinrich, K.; Lazar, M.; Priske, M. High performance organic solvent nanofiltration membranes: Development and thorough testing of thin film composite membranes made of polymers of intrinsic microporosity (PIMs). J. Membr. Sci. 2012, 401402, 222−231. (16) Gorgojo, P.; Karan, S.; Wong, H. C.; Jimenez-Solomon, M. F.; Cabral, J. T.; Livingston, A. G. Ultrathin polymer films with intrinsic microporosity: anomalous solvent permeation and high flux membranes. Adv. Funct. Mater. 2014, 24, 4729−4737. (17) Ong, Y. H.; Shi, G. M.; Le, N. L.; Tang, Y. P.; Zuo, J.; Nunes, S.; Chung, T. S. Recent membrane development for pervaporation processes. Prog. Polym. Sci. 2016, 57, 1−31. (18) Lau, C. H.; Nguyen, P. T.; Hill, M. R.; Thornton, A. W.; Konstas, K.; Doherty, C. M.; Mulder, R. J.; Bourgeois, L.; Liu, A. C. Y.; Sprouster, D. J.; Sullivan, J. P.; Bastow, T. J.; Hill, A. J.; Gin, D. L.; Noble, R. D. Ending aging in super glassy polymer membranes. Angew. Chem., Int. Ed. 2014, 53, 5322−5326. (19) Lau, C. H.; Mulet, X.; Konstas, K.; Doherty, C. M.; Sani, M.-A.; Separovic, F.; Hill, M. R.; Wood, C. D. Hypercrosslinked additives for ageless gas-separation membranes. Angew. Chem., Int. Ed. 2016, 55, 1998−2001. (20) Horn, N. R.; Paul, D. R. Carbon dioxide plasticization and conditioning effects in thick vs. thin glassy polymer films. Polymer 2011, 52, 1619−1627. (21) Lau, C. H.; Konstas, K.; Thornton, A. W.; Liu, A. C. Y.; Mudie, S.; Kennedy, D. F.; Howard, S. C.; Hill, A. J.; Hill, M. R. Gasseparation membranes loaded with porous aromatic frameworks that improve with age. Angew. Chem., Int. Ed. 2015, 54, 2669−2673.

(22) Ben, T.; Ren, H.; Ma, S.; Cao, D.; Lan, J.; Jing, X.; Wang, W.; Xu, J.; Deng, F.; Simmons, J. M.; Qiu, S.; Zhu, G. Targeted synthesis of a porous aromatic framework with high stability and exceptionally high surface area. Angew. Chem. 2009, 121, 9621−9624. (23) Tao, S. J. Positronium annihilation in molecular substances. J. Chem. Phys. 1972, 56, 5499−5510. (24) Eldrup, M.; Lightbody, D.; Sherwood, J. N. The temperature dependence of positron lifetimes in solid pivalic acid. Chem. Phys. 1981, 63, 51−58. (25) Dull, T. L.; Frieze, W. E.; Gidley, D. W.; Sun, J. N.; Yee, A. F. Determination of pore size in mesoporous thin films from the annihilation lifetime of positronium. J. Phys. Chem. B 2001, 105, 4657−4662. (26) Zheng, J.; Lennon, E. M.; Tsao, H.-K.; Sheng, Y.-J.; Jiang, S. Transport of a liquid water and methanol mixture through carbon nanotubes under a chemical potential gradient. J. Chem. Phys. 2005, 122, 214702. (27) Ahmed, A.; Xie, Z.; Konstas, K.; Babarao, R.; Todd, B. D.; Hill, M. R.; Thornton, A. W. Porous aromatic frameworks impregnated with fullerenes for enhanced methanol/water separation. Langmuir 2014, 30, 14621−14630. (28) Hill, R. J. Reverse-selective diffusion in nanocomposite membranes. Phys. Rev. Lett. 2006, 96, 216001. (29) Vandezande, P.; Gevers, L. E. M.; Paul, J. S.; Vankelecom, I. F. J.; Jacobs, P. A. High throughput screening for rapid development of membranes and membrane processes. J. Membr. Sci. 2005, 250, 305− 310. (30) Kuiper, S.; van Rijn, C. J. M.; Nijdam, W.; Elwenspoek, M. C. Development and applications of very high flux microfiltration membranes. J. Membr. Sci. 1998, 150, 1−8. (31) Wu, H.; Gong, Q.; Olson, D. H.; Li, J. Commensurate adsorption of hydrocarbons and alcohols in microporous metal organic frameworks. Chem. Rev. 2012, 112, 836−868. (32) Hill, A. J.; Pas, S. J.; Bastow, T. J.; Burgar, M. I.; Nagai, K.; Toy, L. G.; Freeman, B. D. Influence of methanol conditioning and physical aging on carbon spin-lattice relaxation times of poly (1-trimethylsilyl1-propyne). J. Membr. Sci. 2004, 243, 37−44. (33) Pope, D. S.; Koros, W. J.; Hopfenberg, H. B. Sorption and dilation of poly (1-(trimethylsilyl)-1-propyne) by carbon dioxide and methane. Macromolecules 1994, 27, 5839−5844. (34) Li, Y.; Wee, L. H.; Volodin, A.; Martens, J. A.; Vankelecom, I. F. J. Polymer supported ZIF-8 membranes prepared via an interfacial synthesis method. Chem. Commun. 2015, 51, 918−920. (35) Shao, L.; Cheng, X. Q.; Wang, Z. X.; Ma, J.; Guo, Z. Tuning the performance of polypyrrole-based solvent-resistant composite nanofiltration membranes by optimizing polymerization conditions and incorporating graphene oxide. J. Membr. Sci. 2014, 452, 82−89. (36) Tsarkov, S.; Khotimskiy, V.; Budd, P. M.; Volkov, V.; Kukushkina, J.; Volkov, A. Solvent nanofiltration through high permeability glassy polymers, Effect of polymer and solute nature. J. Membr. Sci. 2012, 423−424, 65−72. (37) Geens, J.; Peeters, K.; Van der Bruggen, B.; Vandecasteele, C. Polymeric nanofiltration of binary water−alcohol mixtures: influence of feed composition and membrane properties on permeability and rejection. J. Membr. Sci. 2005, 255, 255−264. (38) Florian, E.; Modesti, M.; Ulbricht, M. Preparation and characterization of novel solvent-resistant nanofiltration composite membranes based on crosslinked polyurethanes. Ind. Eng. Chem. Res. 2007, 46, 4891−4899.

14408


Hyper-Cross-Linked Additives that Impede Aging and Enhance

Recommend Documents