Article pubs.acs.org/est
Layer-by-Layer Assembly of Aquaporin Z‑Incorporated Biomimetic Membranes for Water Purification Miaoqi Wang,† Zhining Wang,*,† Xida Wang,‡ Shuzheng Wang,† Wande Ding,† and Congjie Gao† †
Key Laboratory of Marine Chemistry Theory and Technology, Ministry of Education, Ocean University of China, Qingdao, Shandong 266100, China ‡ Tianjin Branch of Baotou Research Institute of Rare Earth, Tianjin 300300, China S Supporting Information *
ABSTRACT: We fabricated a biomimetic nanofiltration (NF) membrane by immobilizing an Aquaporin Z (AqpZ)-incorporated supported lipid bilayer (SLB) on a layer-by-layer (LbL) complex polyelectrolyte membrane to achieve excellent permeability and salt rejection with a high stability. The polyelectrolyte membranes were prepared by LbL assembly of poly(ethylenimine) (PEI) with positive charges and poly(sodium 4-styrenesulfonate) (PSS) with negative charges alternately on a porous hydrolyzed polyacrylonitrile (H-PAN) substrate. AqpZ-incorporated 1,2-dioleloyl-sn-glycero-3-phosphocholine (DOPC)/1,2-dioleoyl-3-trimethylammo-nium-propane (chloride salt) (DOTAP) vesicles with positive charges were deposited on the HPAN/PEI/PSS polyelectrolytes membrane surface. The resulting biomimetic membrane exhibited a high flux of 22 L·m−2·h−1 (LMH), excellent MgCl2 rejection of ∼97% and NaCl rejection of ∼75% under an operation pressure of 0.4 MPa. Due to the attractive electrostatic interaction between SLB and the polyelectrolyte membrane, the biomimetic membrane showed satisfactory stability and durability as well as stable NF flux and rejection for at least 36 h. In addition, the AqpZ-containing biomimetic membrane was immersed in a 0.24 mM (critical micellar concentration, CMC) Triton X-100 solution for 5 min. The flux and rejection were slightly influenced by the Triton X-100 treatment. The current investigation demonstrated that the AqpZ-incorporated biomimetic membranes fabricated by the LbL method led to excellent separation performances and robust structures that withstand a high operation pressure for a relatively long time.
■
INTRODUCTION With growing concern for the scarcity of freshwater, membrane technology for water purification and desalination has been consistently developed to increase the water supply through more efficient ways.1 Ideally, a high-performance membrane should exhibit extraordinary permeability and excellent salt rejection with the minimal driving force. The water channel protein, aquaporin (AQP), offers an exciting opportunity to assemble such a highly selective permeable and energy-efficient biomimetic membrane.2−5 AQP is composed of an hourglass structure with its narrowest constriction being 2.8 Å, which prevents solute transport and permits rapid water diffusion.6,7 Several types of biomimetic membranes composed of AQP and phospholipids or copolymers have been designed and fabricated with outstanding desalination performance.8−22 In previous studies, the AQP-incorporated supported lipid bilayer (SLB) membranes prepared via the vesicle rupture method have attracted substantial interest.8−10 This method is a facile and effective approach for preparing AQP-incorporated biomimetic membranes by mimicking the structure of cell membranes (i.e., embedding AQP in SLB membranes). However, low stability is an inherent drawback of the SLB biomimetic membranes. Another design involves mechanically © 2015 American Chemical Society
robust membranes prepared by AQP laden vesicles (proteoliposomes or proteopolymersomes) immobilized in a dense polymer layer.16−22 This type of biomimetic membrane exhibits satisfactory stability and separation performance. However, a large amount of chemicals and detergents can adversely influence the function and reconstitution efficiency of AQP.18,20 In contrast, the SLB biomimetic membrane is expected to exhibit higher water permeability due to its better biocompatibility, and water molecules only need to penetrate the membrane once, whereas the molecules must penetrate twice in the immobilized AQP laden vesicles. The major challenges with the SLB biomimetic membrane involve the fragile structure of the ultrathin selective layer and the defects formed during the vesicles rupture process. To overcome these obstacles, an electrostatic layer-by-layer (LbL) assembly has been proposed as a useful method for preparing a robust Aquaporin Z (AqpZ)-incorporated SLB membrane.23−25 Proteoliposomes covered with positively charged poly-L-lysine Received: Revised: Accepted: Published: 3761
July 2, 2014 January 23, 2015 March 2, 2015 March 2, 2015 DOI: 10.1021/es5056337 Environ. Sci. Technol. 2015, 49, 3761−3768
Article
Environmental Science & Technology
on the H-PAN substrate. The duration time of each deposition was 30 min, and each adsorption was followed by a rinse with DI water for 5 min to remove the redundant polyelectrolytes. This procedure was repeated such that multilayer polyelectrolyte membranes with a specific bilayer number were fabricated. The prepared polyelectrolytes membranes were denoted as LbLN, where N represents the number of polyelectrolyte bilayers. Preparation of Liposomes and Proteoliposomes. A DOPC or DOPC/DOTAP mixture with a molar ratio of DOPC/DOTAP = 4/1 was used to prepare the liposome samples. The samples were prepared by using the film rehydration method according to a previously reported protocol.30 The resulting vesicle solution was extruded 11 times with 100 nm polycarbonate membranes through an extruder system (Avanti Polar Lipids, Alabaster, AL) to obtain unilaminar vesicles with a final concentration of 0.1 mg/mL. A certain amount of the AqpZ solution was added to the liposome solution containing 1% OG with different protein to lipid mass ratios (P/L = 1/400, 1/200, 1/100, 1/50 w/w). The solution was dialyzed in a 6−8 kDa molecular weight cutoff (MWCO) dialysis bag (Spectra Por) against 1 L of phosphate buffer for 3 days at room temperature. The buffer was refreshed every day. After dialysis, the proteoliposome solution was extruded through the polycarbonate membrane with a mean pore size of 100 nm as previously mentioned prior to further use. Formation of SLB on the LbL Polyelectrolytes Membranes. Fifteen milliliters of the 0.1 mg/mL liposome or proteoliposome solution was spread onto the LbL polyelectrolyte membrane and incubated at room temperature for 4 h in a homemade membrane cell. After fusion, the biomimetic membrane was rinsed with DI water several times to remove the residual vesicles. Vesicles Characterizations. A stopped-flow spectrometer (Applied Photophysics Ltd., UK) was used to characterize the permeability of the liposomes and proteoliposomes. The samples were rapidly mixed with a hyperosmolar sucrose PBS buffer. The osmolarity difference (Δosm) was measured to be 290 mosm/L with an osmometer (Osmomat 030, Gonotec, Germany). The variations in the vesicle size were detected and recorded by light scattering at an emission wavelength of 577 nm in a stopped-flow apparatus. Each sample was tested at least three times, and the repeatability was very good. The measuring temperature was 22 ± 0.1 °C. An exponential rise equation (eq 1) was employed to fit the acquired data curves.
have been immobilized on a negatively charged polyelectrolyte membrane by electrostatic interaction to form a robust AQPladen mixed matrix membrane.19 In this work, positively charged proteoliposomes were deposited on a negatively charged membrane, which was assembled with polyelectrolytes via LbL deposition to form a robust and continuous SLB biomimetic membrane. As shown in Figure 1, poly(ethylenimine) (PEI) was deposited on top of
Figure 1. Schematic presentation of the formation procedures for the AqpZ-incorporated LbL membrane.
a hydrolyzed polyacrylonitrile (H-PAN) membrane to form a polycation layer. Then, polystyrenesulfonate (PSS) was deposited to form the polyanion layer. Next, 1,2-dioleloyl-snglycero-3-phosphocholine (DOPC)/1,2-dioleoyl-3-trimethylammo-nium-propane (chloride salt) (DOTAP) proteoliposome, which is positively charged, was deposited onto the LbL membrane to form a SLB via vesicle rupture. The aim of our work was to improve the stability and mechanical strength of the AQP-embedded SLB membrane. Our method provided a new procedure for the fabrication of a robust biomimetic membrane with a high water purification performance.
■
MATERIALS AND METHODS Materials and Chemicals. Polyacrylonitrile (PAN, Mw ∼50000) was kindly supplied by the Shanghai Jingshan Petrochemical Company. Polyethylenimine (PEI, M w ∼750,000), poly(sodium 4-styrene-sulfonate) (PSS, M w ∼70,000), 1-n-octyl-β-D-glucopyranoside (OG, purity >99%), and Triton X-100 (purity >99%) were purchased from SigmaAldrich (USA). 1,2-Dioleloyl-sn-glycero-3-phosphocholine (DOPC, purity >99%) and 1,2-dioleoyl-3-trimethylammonium-propane (chloride salt) (DOTAP) were obtained from Avanti Polar Lipids (USA). AqpZ was kindly supplied by Prof. Zhinan Xu from Zhejiang University (China).26−29 The other materials were of the highest purity available. Milli-Q water (Millipore, integrated ultrapure water system) with a resistivity of 18.2 MΩ·cm was used. Preparation of Porous PAN Substrates. The PAN casting solution was prepared by dissolving PAN and LiCl in N,N-dimethylformamide (DMF) at a mass ratio of 18:2:80. Porous PAN substrate membranes were fabricated using a phase inversion method. The resulting PAN membrane was immersed in a 1.5 M NaOH solution at 45 °C for 1.5 h.24 The alkali solution treated PAN membrane was denoted as H-PAN. LbL Assembly of Polyelectrolytes. The LbL polyelectrolyte membrane was prepared by alternatingly depositing a 1 wt % PEI aqueous solution and a 1 wt % PSS aqueous solution
Y = A exp( −kt )
(1)
where Y is the intensity of the signal, t is the recording time, A is a constant, and k is the initial rate constant. The osmotic water permeability was calculated from eq 2: Pf = k /((S /V0) × VW × Δosm)
(2)
where S is the vesicle surface area, V0 is the initial volume of vesicles, VW is the partial molar volume of water (0.018 L·mol−1), Δosm is the osmolarity difference that drives the shrinkage of the vesicles, and Pf is the osmotic water permeability. Zetasizer Nano ZS90 (Malvern, UK) was employed to characterize the vesicle size distribution and surface charge (zeta-potential, δ). A 633 nm laser source with a fixed detector angle of 90° was used in the dynamic mode. All of the 3762
DOI: 10.1021/es5056337 Environ. Sci. Technol. 2015, 49, 3761−3768
Article
Environmental Science & Technology measurements were carried out at 22 ± 0.1 °C and performed in triplicate. Membranes Characterizations. The membranes were analyzed by a Fourier transform infrared (FTIR) spectrometer (Tensor 27, Bruker, Germany) using the attenuated total reflection (ATR) technique to detect the functionalized groups on the membranes. The samples were dried at 40 °C in vacuum for 24 h prior to FTIR measurement to minimize the influence of water.8 Scanning electron microscopy (SEM, S-4800, Hitachi, Japan) was used to analyze the morphology and structure of the original and modified membranes. The membranes were air-dried overnight prior to analysis. The dry membrane samples were frozen in liquid nitrogen and subsequently cracked to obtain the cross sections. The samples were mounted on sample stages using double-sided conductive tape and sputter-coated with gold for 50 s. Atomic force microscopy (AFM, Nanoscope V MultiMode, Veeco, USA) was applied to investigate the surface morphology and roughness of the membranes. The AFM images were obtained by in situ measurements. Tapping mode measurements in liquid were performed using NP-S cantilevers (short lever, nominal spring constant 0.06 N·m−1). The surface roughness was defined as the root-mean-squared height (RMS) over a membrane surface area of 5 × 5 μm2 and obtained from AFM images using the instrument software. The zeta potential values of the membranes were analyzed on the basis of the streaming potential measurement by an electrokinetic analyzer (Anton Paar SurPASS, Austria) with 1 mM KCl as the electrolyte solution at a pH of 7.5. Nanofiltration (NF) Performances. The permeation flux and salt rejection were measured using a homemade cross-flow NF apparatus under a transmembrane pressure of 0.4 MPa. The flow rate was 40 L/h, and the velocity was 20 cm/s. The effective area of the membrane was 19.56 cm2. 0.5 g/L NaCl and 0.5 g/L MgCl2 were used as the feed solutions. The permeation flux (JV) was determined at least 1 h after the start of the NF tests to allow for stabilization. JV was calculated by eq 3: JV = Q /At
Figure 2. Stopped-flow results for the DOPC and DOPC/DOTAP liposomes and proteoliposomes. (A) Normalized light scattering of the stopped-flow kinetics curves and (B) water permeability coefficient.
liposomes, which provide more water channels to enhance the permeability. The permeability of AqpZ in the DOPC/DOTAP liposome (P/L = 1/50) was calculated to be 32.4 × 10−14 cm3· s−1 per unit, which agrees well with the previously reported results.21,31,32 The mean vesicle diameters and the zeta potential of the liposomes and proteoliposomes have been measured using a dynamic light scattering (DLS) apparatus. As shown in Table 1, the vesicle samples exhibit a narrow size range from 114 to 122 nm and low polydispersity indices (
