As(III) Removal from Drinking Water by Carbon Nanotube Membranes

May 24, 2018 - As(III) Removal from Drinking Water by Carbon Nanotube Membranes with Magnetron-Sputtered Copper: Performance and Mechanisms...
0 downloads 0 Views 8MB Size
Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES

Energy, Environmental, and Catalysis Applications

As(III) Removal from Drinking Water by Carbon Nanotube Membranes with Magnetron-Sputtered Copper: Performance and Mechanisms Hongyan Luan, Quan Zhang, Guo-An Cheng, and Haiou Huang ACS Appl. Mater. Interfaces, Just Accepted Manuscript • Publication Date (Web): 24 May 2018 Downloaded from http://pubs.acs.org on May 24, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

As(III) Removal from Drinking Water by Carbon Nanotube Membranes with Magnetron-Sputtered Copper: Performance and Mechanisms

Hongyan Luan†, Quan Zhang‡, Guo-an Cheng‡, Haiou Huang†* †

State Key Joint Laboratory of Environmental Simulation and Pollution Control,

School of Environment, Beijing Normal University, No. 19, Xinjiekouwai Street, Beijing 100875, China ‡

Key Laboratory of Beam Technology and Material Modification of the Ministry of

Education, College of Nuclear Science and Technology, Beijing Normal University, No. 19, Xinjiekouwai Street, Beijing 100875, China * Department of Environmental Health Sciences, Bloomberg School of Public Health, The John Hopkins University, 615 North Wolfe Street, MD 21205, USA

KEYWORDS: arsenite; carbon nanotube membranes; Cu nanoparticles; drinking water treatment; magnetron sputtering

ABSTRACT: Current approaches for functionalizing carbon nanotubes (CNT) often utilize harsh chemical conditions and the resulting harmful wastes can cause various environmental and health concerns. In this study, magnetron-sputtering technique is 1

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

facilely employed to functionalize CNT membranes by depositing Cu onto pre-made CNT membranes without using any chemical treatment. A comparative evaluation of the substrate polymeric membrane (MCE), MCE sputtered with copper (Cu/MCE), the pristine CNT membrane (CNT), and CNT membrane sputtered with Cu (Cu/CNT) shows that Cu/CNT possesses mechanically stable structures and similar membrane permeability as MCE. More importantly, Cu/CNT outperforms other membranes with high As(III) removal efficiencies of above 90%, as compared to less than 10% by MCE and CNT, and 75% by Cu/MCE from water. The performance of Cu/CNT membranes for As(III) removal is also investigated as a function of ionic strength, sputtering time, co-existing ions, solution pH, and the reusability. Further characterizations of As speciation in the filtrate and on Cu/CNT reveal that arsenite removal by Cu/CNT possibly began with Cu-catalyzed oxidation of arsenite to arsenate, followed by adsorptive filtration of arsenate by the membrane. Overall, this study demonstrates that magnetron sputtering is a promising greener technology for the productions of metal-CNT composite membranes for environmental applications.

1. INTRODUCTION The issue of water scarcity and pollution has seriously threatened human life, and thus, novel water treatment technologies are in immediate need for sustainable water around the world. In the past three decades, membrane technologies have gained increased attention in the field of water treatment because of its flexibility, efficiency and environmental friendly.1 Among current-generation membranes, low-pressure membranes (LPMs) have attracted great interests owing to their low operational transmembrane pressures (TMP, which are typically less than 0.5 bar), as compared to those of nanofiltration (NF) or reverse osmosis (RO) membranes.2 As a result, LPMs consume much less energy than NF/RO at similar water treatment capacities.

2

ACS Paragon Plus Environment

Page 2 of 46

Page 3 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

However, traditional LPMs possess relatively large pore size, ranging from 0.01 µm to a few microns, and therefore, are only capable of sieving out particulate matter in water while being ineffective for dissolved substances including arsenite. Consequently, filtration of dissolved contaminants by LPMs has been an important effort made by scientists and engineers in the past decade.3, 4 In this regard, carbon nanotubes (CNTs) have been increasingly studied for the development of novel microfiltration (MF) membranes with pore size ranging from 0.1 to a few microns, owing to their high water permeability, high sorption capacities, as well as outstanding physical and chemical properties.5 Suitable techniques are currently being investigated in order to functionalize CNTs to enhance the water permeability and contaminant selectivity of CNT membranes.6, 7 In recent years, great interest in metal-CNT composites has risen in scientific community because of the versatile performance demonstrated by these composites in different applications.8, 9 However, the preparation of metal-CNT composites often necessitates the use of harsh chemical conditions. In addition to the chemical hazard incurred during the reaction process, chemical wastes produced after chemical functionalization may also pollute the environment if not being treated properly. Therefore, environmentally friendly approaches are being sought to prepare metal-CNT composites suitable for the fabrication of LPMs. Magnetron sputtering is a physical vapor deposition technique. In the sputtering process, a target (or cathode) plate is bombarded by energetic ions generated in a glow discharge plasma which is situated in front of the target. The bombardment process removes from the target and then deposits them as a thin film on the substrate.10 This technique possesses many advantages over chemical coating method, including ease of operation, high efficiency and environmental friendliness. Also, the coating layers formed by magnetron sputtering are more uniform, compact and durable than those formed by other physical coating methods.11 Moreover, a variety of metallic materials 3

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

can be used as the target sources for thin-film deposition. This technique has been widely used for the manufacturing of wear-proof and lubricating thin films, as well as optics and electronic semiconductors for various industries.12 The applicability of magnetron sputtering to nanomaterial fabrication has been reported recently. For example, Lv et al. modified a nanoscale bacterial-derived cellulose with copper nanoparticles by magnetron sputtering, the products exhibited excellent electromagnetic shielding, as well as thermal, conduction, and mechanical properties.13 Using the same technique, Pham et al. deposited Pt on graphene-CNT hybrids for the proton exchanging membrane fuel cells.14 Besides, magnetron sputtering was applied to functionalize the surface of polyester fiber with silver nanoparticles.15 Despite the aforementioned merits, this technique has rarely been studied for the preparation of nanocomposite membranes for water treatment. Therefore, the objectives of this study were: 1) to prepare Cu/CNT composite membranes by magnetron sputtering, 2) to examine the performance of the prepared membranes for the removal of As(III) from drinking water, and 3) to explore the mechanisms of As(III) filtration by the Cu/CNT membranes. Arsenite was chosen as the target contaminant in this study because it is a highly toxic substance and excessive ingestion of arsenite via drinking water can cause skin lesions, cancer and other serious health problems.16 In consequence, we prepare different CNT membranes by a vacuum-filtration method and further functionalize the membranes with Cu nanoparticles by magnetron sputtering. In order to evaluate the performance of different Cu-sputtered membranes, we investigate the effects of ionic strength, sputtering time, competing ions (sulfate and phosphate), initial solution pH on As(III) removal and the reusability of the Cu/CNT membrane. Morphological and chemical properties of the Cu/CNT membranes are also determined by optoelectronic techniques to reveal potential mechanisms for As(III) removal. The results obtained in the study show that 4

ACS Paragon Plus Environment

Page 4 of 46

Page 5 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

magnetron-sputtered Cu/CNT membrane indeed possesses a promising potential for As(III) removal from drinking water. This finding provides important insights into the applicability of magnetron sputtering technique for the preparation of metal-CNT composite membranes for water treatment.

2. EXPERIMENTAL SECTION 2.1. Materials and Chemical Reagents. A carboxylated multi-walled carbon nanotube (MWCNT-COOH) was used as received in this study for the preparation of CNT membranes due to its good dispersity in water. This CNT material was purchased from the Beijing Boyu Technology Corporation of High-tech New Materials, China. According to the manufacturer, this CNT sample had an outer diameter of < 8 nm, lengths between 10 - 30 µm, a specific surface area (SSA) of 500 m2 g-1, and a purity of > 95%. Furthermore, reagent-grade of NaAsO2 and NaCl were bought from Sinopharm Chemical Reagent Co., Ltd, China, and dissolved into ultrapure water (18.2 MΩ cm at 25 °C) to prepare stock solutions at desired concentrations. We also used the hydrophilic mixed cellulose ester (MCE) membrane (GSWP04700, Millipore, USA) as the supporting layer for the preparation of composite membranes described below. This membrane possessed a flat-sheet configuration and a nominal pore size of 0.22 µm.

2.2. Preparation of Composite Membranes. The CNT-composite membranes was prepared by a vacuum filtration method developed by Wu et al.17 Specifically, 12 mg of MWCNT-COOH powder were dispersed into 30 mL of ultrapure water and sonicated with a probe sonicator (Ultrasonic processor FS-250N, Shanghai Sonxi Ultrasonic Instrument Co., Ltd, China) at 150W for 10 minutes. Afterwards, the MWCNT-COOH dispersion was quickly poured into a funnel mounted in a vacuum 5

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

filtration setup and filtered at a vacuum pressure of 0.04 MPa through a MCE membrane placed at the bottom of the funnel. Finally, the CNT membrane was air-dried at an ambient temperature of 25 °C. A magnetron sputtering deposition apparatus manufactured by College of Nuclear Science and Technology, Beijing Normal University was used for the deposition of copper onto the aforementioned CNT membrane. For comparison, a virgin MCE membrane was also modified with Cu under similar conditions. For magnetron sputtering of Cu, a CNT membrane or a MCE membrane was placed inside a vacuum chamber inside the magnetron-sputtering apparatus. During the sputtering process, the chamber was vacuumized to below 10-4 Pa. After that, the membrane was sputtered from a copper target (99.999% purity) at a direct current of 99 mA with 10 sccm Ar (99.995%) at a controlled pressure of 1.5 Pa. All membranes used in the following experiments were sputtered for 10 min to allow uniform Cu coating of Cu based upon preliminary tests. Given a preset deposition rate of 0.2 nm/s for Cu, the thickness of Cu coating on the membrane surfaces was expected to be approximately 120 nm and the amount of Cu on the composite membrane was approximately 0.524 mg calculated based upon the density and volume of copper. To explore the effect of Cu amount on As(III) removal, different sputtering durations (5min, 15min) were also applied during the functionalizing process.

2.3. Characterizations of the Composite Membranes. Stabilities and flexibilities of CNT and Cu/CNT were evaluated by external forces created by bending. Surface morphologies of the virgin and exhausted membranes were obtained by using a cold cathode field emission scanning electron microscope (FESEM) (HITACHI S-4800) with a 1.0 nm resolution at 10 keV. Energy-dispersive spectroscopy (EDS) equipped in FESEM was used to determine the element compositions of the as-prepared Cu/CNT membranes. Surface chemical composition 6

ACS Paragon Plus Environment

Page 6 of 46

Page 7 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

and elemental speciation of the membranes were measured by an X-ray photoelectron spectroscope (XPS, ESCALAB 250Xi, USA).

2.4. Membrane Filtration Experiments. The arsenite filtration performance of the virgin MCE membrane, the Cu-sputtered MCE or Cu/MCE membrane, the CNT membrane, and the Cu-sputtered CNT or Cu/CNT membrane was assessed comparatively based upon their arsenite removal efficiencies under different feed water quality conditions. The bench-scale filtration system used in this study consisted of a dual-channel syringe pump (RSP02-B, RISTRON, China), two 50-mL polystyrene syringes, and two polypropylene syringe filter holders (Whatman, 420200, UK). The effective diameter of the filter holders was 25 mm, resulting in an effective membrane area of 4.9×10-4 m2 for each filter. During each filtration experiment, the feed solution spiked with arsenite was filtered through the parallel membranes at a constant permeate flux of 6.7 L m-2 h-1. A permeate sample was collected for every 40 mL of filtrate, which was equivalent to a permeate throughput of 81.6 L m-2. The temperature for all filtration experiments was controlled at 25 ± 1 °C.The As removal efficiency was calculated by the following equation. As removal (%) = (C0 - Ct)/C0 × 100% where C0 is the feed concentration of As, Ct is the concentration of As in the permeate sample. To determine the variation of As species in the permeates, we collected the samples of each 5 mL filtered solution, and the total filtration volume was 40 mL. 2.4.1. Effect of Cu Sputtering on As(III) Removal. A series of filtration experiments were performed to reveal the effect of Cu sputtering on As(III) removal. In this study, bare membranes without Cu sputtering (MCE and MWCNT-COOH) and the sputtered membranes of Cu/MCE and Cu/MWCNT-COOH were employed to illustrate the influence of Cu sputtering on As(III) removal. Not only that, influence of sputtering 7

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

time on As(III) removal was also studied with the Cu/MWCNT-COOH membrane sputtered for 5 and 15 min, respectively. Furthermore, different feed solutions were prepared in order to systematically investigate the effect of solution chemistry on arsenite removal. For this purpose, the ionic strength of the feed water was varied at 0, 0.01 and 0.1 M, by adding NaCl at different molar concentrations at a fixed pH of 7.0. All feed solutions were spiked with 100 µg L-1 of As(III) in the form of NaAsO2. 2.4.2. Sorption Kinetics. The sorption kinetic tests were conducted at an elevated As(III) concentration of approximately 4.74 mg L-1 with the Cu/CNT membrane. The pH of the feed solution was adjusted to 7.0. During the filtration process, samples were taken at 100 min intervals, and the arsenic concentration in the permeate was measured by ICP-AES analysis. 2.4.3. Effects of Competing Anions. The impacts of competing anions on As(III) removal efficiency were carried out in the presence of SO42-, HPO42- with three different levels (0, 0.2 and 5 mM). The As(III) concentration was still 100 ppb, and the pHs were adjusted to 7.0. 2.4.4. Regeneration and Reusability of Cu/CNT Membrane. To evaluate the reusability of Cu/CNT membrane, a Cu/MWCNT-COOH membrane was employed to filter the As-bearing water with concentration of 100 ppb, and the experiment was conducted at neutral pH. The membrane was regenerated by 10 mL of 0.001 M NaOH after each 40 mL permeate was filtered. Five filtration-regeneration cycles were conducted to evaluate the reusability of the Cu/CNT membrane. 2.4.5. Effect of Solution pH. Effects of initial solution pHs on arsenic removal were investigated during the filtration experiments, the pHs were adjusted to 6.0, 7.0 and 8.0 with 0.1 M HCl and NaOH solution. The concentrations and speciation of As in the permeate were determined in further analysis. 2.4.6. Measurement of Cyclic Voltammetry. Cyclic voltammetry (CV) was carried 8

ACS Paragon Plus Environment

Page 8 of 46

Page 9 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

out to evaluate the occurrence of oxidation reactions on the studied membranes immersed in a testing solution containing 0.01 M NaCl and 1.0 mg L-1 of As(III). The pH of the solution was adjusted to 7.0. CV tests were conducted with a three-electrode system. Before each measurement, a membrane was installed onto an electrochemical station (CHI660E, CH Instruments, Inc. USA) as the working electrode, with a platinum electrode acting as the counter electrode and an Ag/AgCl electrode worked as the reference electrode. During the CV analysis, the system was operated at a scanning rate of 2 mV s-1 with a potential sweep of -0.6 V to 0.6 V. The measurement was repeated for 20 cycles to ensure reproducibility of the results.

2.5.

Analytical methods.

The concentrations of As and Cu in the permeates were determined by an inductively coupled plasma-atomic emission spectrometry system (ICP-AES) (SPECTRO ARCOS EOP, SPECTRO Analytical Instruments GmbH/Spectro) at detection limits of 1.2 µg L-1 and 0.5 µg L-1, respectively. Arsenic species in the feed solution and permeate samples were also determined in selected experiments by a high performance liquid chromatography (HPLC, Agilent 1100) and ion coupled plasma-mass spectrometry (ICP-MS, Agilent 7500ce) hyphenated system (Agilent Technologies, Palo Alto, CA, USA). The mobile phase of the HPLC consisted of 15 mM of (NH4)2HPO4 and was pumped through the column at 1.0 mL min-1. The resultant retention time of As(III) peak was 2.635 min and As(V) was 10.663 min as determined by As(III) and As(V) standard solutions.

3. RESULTS AND DISCUSSION 3.1. Characteristics of magnetron-sputtered composite membranes. As shown in Figure S1, vacuum filtration of the CNT dispersion forms a stable CNT layer on 9

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the MCE membrane. The resultant CNT membrane is stable and can be bent from any angles (Figure S1a and b). After magnetron sputtering of Cu onto the MCE membrane or the CNT membrane, the resulting Cu/MCE or Cu/CNT membrane also shows good flexibility and no sign of CNT or Cu dropping from the supporting MCE membranes (Figure S1c and d). Moreover, the Cu/MCE membrane and the Cu/CNT membrane exhibit brass color, i.e., the native color of elemental copper, as compared to the black color of the CNT membrane. These results indicate that all modified membranes possess structures that are stable with respect to the external forces created by bending, a propensity that is beneficial from the standpoint of sustainable membrane water treatment. Furthermore, the virgin MCE membrane has a bi-continuous porous structure typical for polymeric MF membranes (Figure 1a). After being directly treated by magnetron sputtering, the polymeric matrix of the MCE membrane is completely covered by a thin layer of Cu (Figure 1b). In comparison, CNT forms clumpy structures on top of the MCE membrane (Figure 1c). After being sputtered with Cu, the clumps of CNT on the membrane are wrapped completely by spheroidal-shaped copper nanoparticles with low resolution of SEM (Figure 1d). In Figure 1e, with high resolution of SEM, the fibrous structure of CNT is clearly observed, forming a porous network on top of the MCE membrane. After sputtering, the fibers of CNT are enveloped with nano-copper whose diameters are less than 100 nm, thus the CNT is not seen in the SEM image (Figure 1f). The morphologies of Cu/MCE and Cu/CNT look different as shown in Figure 1b and f. The Cu film deposited on the MCE membrane appears to be smoother than that on the CNT membrane. This difference is possibly attributable to the flatter surface and larger structure size of the MCE membrane, which led to the formation of smoother and continuous Cu film deposited on the MCE fibers. Comparatively, the CNT membrane possesses discrete CNT aggregates, which facilitated the formation of particulate deposits on membrane surfaces. The morphologies of the Cu/CNT membranes sputtered with Cu for 5 and 10

ACS Paragon Plus Environment

Page 10 of 46

Page 11 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

15 min are shown in Figure 1g and h. The CNT fibers are completely coated with Cu nanoparticles despite the use of different sputtering time. Moreover, SEM-EDS images show that the distribution of Cu nanoparticles on different Cu/CNT membranes is uniform and identical (Figure S2). XPS analysis is also performed to determine the surface functionality of the studied membranes. As shown in Figure 2a and b, the surfaces of the Cu/MCE membrane and the Cu/CNT membrane contain similar atomic percentages of copper (ca. 40%), oxygen (ca. 30%), and carbon (20%), as well as trace amounts of nitrogen. According to Espino et al., the binding energies of metallic Cu, Cu2O, and CuO are approximately 932.7, 932.5 and 933.6 eV, respectively.18 Also, CuO has four Cu(II) peaks which can be distinguished directly while Cu(0) and Cu(I) have two peaks.19 These findings are consistent with our results. As shown in Figure 2c and d, the Cu2p spectra of the sputtered membranes have two distinctive peaks corresponding to Cu2p3/2 and Cu2p1/2, respectively, as well as two small satellite peaks. According to these XPS results, the chemical states of the sputtered copper on the membranes are primarily Cu(0), with a small amount of Cu(I) or Cu(II).

11

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

12

ACS Paragon Plus Environment

Page 12 of 46

Page 13 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 1. Scanning electron micrograms of the top surfaces of (a) the MCE membrane with high resolution (×50,000), scale bar = 500 nm; (b) the Cu/MCE membrane; (c) the CNT membrane, with low resolution (×5,000), scale bar = 5 µm; (d) the Cu/CNT membrane with low resolution (×5,000); (e) the CNT membrane with high resolution (×50,000), scale bar = 500 nm; (f) the CNT membrane sputtered with Cu for 10 min with high resolution (×50,000); (g) the CNT membrane sputtered with Cu for 5 min with high resolution (×50,000); (h) the CNT membrane sputtered with Cu for 15 min with high resolution (×50,000).

Figure 2. XPS spectra of (a) the Cu/MCE membrane and (b) the Cu/CNT membrane, and high-resolution Cu2p spectra of (c) the Cu/MCE membrane and (d) the Cu/CNT 13

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

membrane.

3.2. Effect of Cu sputtering on As(III) removal. As a benchmarking condition, less than 10% of As(III) spiked in ultrapure water is removed by the virgin MCE membrane or the CNT membrane (Figure 3a). Comparatively, membranes sputtered with copper exhibit noticeably improved arsenite removals. For the MCE membrane, Cu sputtering increases As(III) removal efficiencies from approximately 4% to 75% (Figure 3b). For the CNT membrane, the arsenic removal efficiencies increase to more than 90% after magnetron sputtering of Cu nanoparticles onto the membrane. These results suggest that the sputtered Cu nanoparticles play a primary role in the arsenite removal by the composite membrane, while the presence of MWCNT-COOH enhance the efficacy of Cu nanoparticles for arsenite removal. On the other hand, different forms of copper coating observed in Figure 1b and f for Cu/MCE and Cu/CNT membranes may affect the removal of arsenite, which needs to be investigated in future study. On the other hand, increase in ionic strength appears to negatively affect arsenite removal by the Cu-sputtered membranes (Figure 3b). When the ionic strengths of the background solution increase, the As(III) removal efficiencies of the Cu/MCE membrane decrease from more than 75% with the ultrapure water to 50% with 0.01 M NaCl (representative of fresh water), and then sharply to approximately 17% with 0.1 M NaCl (representative of moderately brackish water). In comparison, the As(III) removal efficiencies of the Cu/CNT membrane remain at 90% as the ionic strength increased to 0.01 M but significantly lost its efficiency when the ionic strength increased to 0.1 M. Regardless of the ionic strength, As removals by the Cu/CNT membrane are always greater than those by the Cu/MCE membrane, suggesting the importance of CNT as a membrane component.

14

ACS Paragon Plus Environment

Page 14 of 46

Page 15 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

The negative effect of ionic strength on As(III) removal is consistent with the finding of a previous study. Payne and Abdel-Fattah observed that when the concentration of KNO3 increased from 0.01 M to 0.1 M, As(III) adsorption by Fe-modified activated carbon declined by 20%.20 This negative effect leads to a speculation that the removal of arsenic by Cu-sputtered membranes involves adsorption of arsenic species onto the surfaces of Cu nanoparticles, owing to the formation of outer-sphere complexes through electrostatic forces.21 The increases in ionic strength inhibit arsenic adsorption due to increased competitive adsorption of anions such as Cl-. Furthermore, effects of Cu amount on arsenic removal are also studied by varying the sputtering time for the Cu/CNT membranes. From Figure 3c, when the sputtering time is 5 min, the removal efficiency of As(III) is 75%, lower than that of the Cu/CNT membrane sputtered for 10 min (Figure 3b). These results can be ascribed to the various amounts of Cu nanoparticles coated on the CNT membranes. The amount of nano-copper increases with the extension of sputtering time. Therefore, coating more nano-copper on the CNT membrane may enhance the removal of As(III). However, when the sputtering time is extended to 15 min, the removal ratio of As(III) decrease to 67% because the excessive nano-copper forms a dense film that is less permeable to As-laden water, resulting in decreased As removal (Figure 3c).

15

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3. Removals of arsenic spiked in the ultrapure water by membranes without Cu-sputtering (a), and spiked in background solutions with different NaCl concentrations by Cu-sputtered membranes (b); effects of sputtering time on As(III) removal with Cu/CNT membranes with background solution of 0.01 M NaCl (c). NaCl concentrations = 0, 0.01 and 0.1 mol L-1, initial As(III) concentration = 100 µg L-1, pH = 7.0, temperature = 25 ± 1 °C.

3.3. Sorption kinetics of the Cu/CNT membrane. Figure 4 displays the sorption kinetics of the Cu/CNT membrane. The initial concentration of arsenic is 4.74 ppm and the filtration run last for 700 min. It is shown that the sorption kinetics of the Cu/CNT membrane is fitted well with the pseudo-first-order kinetic model with high correlation coefficients (R2 = 0.9886, Figure 4a), instead of fitting well with the pseudo-second-order kinetic model of other adsorbents for the arsenic removal,22,23 owing to the differences between dynamic filtration and static adsorption. The 16

ACS Paragon Plus Environment

Page 16 of 46

Page 17 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

corresponding adsorption capacity of the Cu/CNT membrane toward As(III) is 45 mg g Cu-1 calculated by the equilibrium concentration of As, higher than other materials summarized in Table 1.

Figure 4. Sorption kinetics of the Cu/CNT membrane for the removal of As(III). (a) pseudo-first-order

kinetic

model;

(b)

pseudo-second-order

kinetic

model.

Experimental conditions: arsenic concentration = 4.74 ppm, pH = 7.0, constant permeate flux = 6.7 L m-2 h-1, and temperature 25 °C.

Table 1. Comparison of As(III) adsorption capacities with various absorbents.

Adsorbents

Sorption capacity

pH

References

(mg g-1) doughnut-like CuO

4.7

4.0

24

TiO2-coated carbon nanotube

13.6

7.0

25

Al2O3/Fe(OH)3

9.0

6.6

26

1.0862

7.0

27

Copper (II) oxide

17

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Copper oxide incorporated

Page 18 of 46

2.61

7.0

28

1.37

7.0

29

zerovalent iron

3.5

7.0

30

Cu/MWCNT-COOH membrane

45

7.0

Present study

mesoporous alumina Copper exchange zeolite-A (CEZ)

3.4. Effects of competing anions. Sulfate and phosphate are widely present in groundwater which may influence the adsorption of arsenic by competing for binding sites on Cu/CNT membranes. In this study, sulfate and phosphate exert different effects on arsenic removal. When the concentration of sulfate is 0.2 mM, the removal efficiency of arsenic decreases from 94% to 74%; when the concentration of sulfate increased to 5 mM, the removal rate drops to 64% (Figure 5). These results suggest that higher concentration of sulfate plays a negative role in the removal of arsenic. More importantly, even when the concentration of phosphate is 0.2 or 5 mM, As removal efficiencies decline dramatically to 43% or 20% (Figure 5), indicating that the presence of phosphate significantly inhibits the removal of As(III) as HPO42competes for the sorption sites on Cu/CNT membrane with arsenic species.

18

ACS Paragon Plus Environment

Page 19 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 5. Effects of sulfate and phosphate on the removal of As(III). Initial As concentration = 100 ppb, pH = 7.0, temperature 25 °C.

3.5. Regeneration and reusability of Cu/CNT membrane. After saturated adsorption of As(III) with the Cu/CNT membrane, a 0.001 M NaOH solution is used to regenerate the exhausted membrane. The NaOH concentration is chosen as the upper alkaline level compatible with the substrate MCE membrane. The results show that after regeneration, the removal ratio of As(III) is 70% at the first filtration stage (Figure 6). For the further filtration-regeneration cycles, the As removal efficiencies decline slightly compared with the first stage (Figure 6). The results indicate that the 0.001 M NaOH solution cannot regenerate the Cu/CNT membrane effectively. Therefore, in the future study, various regenerating solutions will be evaluated for desorption of arsenic from Cu/CNT membranes.

19

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure

6.

As(III)

removal

with

Cu/CNT

membrane

Page 20 of 46

for

different

filtration-regeneration cycles. Initial As(III) concentration = 100 ppb; regenerated by 0.001 M NaOH; constant permeate flux = 6.7 L m-2 h-1; permeate volume of each sample = 40 mL.

3.6. XPS analysis of the Cu/CNT membrane. In order to explore the mechanism of As(III) removal, XPS analysis is conducted with the Cu/CNT membrane after filtration of an As(III) solution and the results are compared to those obtained with a similar membrane before arsenic filtration. The feed solution contains approximately 100 µg L-1 of As(III) in the background solution of 0.01 M NaCl. As shown in Figure 7a, the atomic percentage of Cu on membrane surfaces decreases from approximately 42% to approximately 21% after arsenic filtration, possibly due to the reaction between arsenite and Cu on membrane surfaces. Furthermore, after As filtration, the Cu2p spectrum has two new distinct and strong Cu2+ satellite peaks (Figure 7b) that do not exist before arsenite filtration (Figure 2d), meaning that the main state of Cu on the composite membrane became Cu(II) which may be due to the reactions between nano-copper and the synthetic water. These results suggest that magnetron-sputtered Cu(0) was oxidized to Cu(II) during the arsenite filtration process. The SEM image of Cu/CNT membrane after treatment with As(III) also 20

ACS Paragon Plus Environment

Page 21 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

confirms that the nano-coppers coated on the CNT membrane participate in the chemical reactions, resulting in minor erosion of Cu nanoparticles (Figure S3). Coincident with the appearance of Cu(II) peaks, high-resolution As3d spectrum of the used Cu/CNT membrane (Figure 7c) also demonstrates that As immobilized on membrane surface possesses two chemical states at the binding energies of 44.3 and 45.2 eV, corresponding to As(III) (44.3 - 44.5 eV) and As(V) (45.2 - 45.6 eV), respectively.31-33 Based upon the relative peak areas, approximately 81.7% of arsenic filtered on the membrane is As(V) oxidized from As(III) during the filtration process, while 18.3% of As(III) is adsorbed directly onto the membrane. This result agrees with the common view that As(V) is less mobile than As(III) in natural or engineered systems.34 Moreover, the increased oxygen content after arsenite filtration (Figure 7a) and shifting of O1s spectrum by approximately 1.1 eV to the higher-energy edge (Figure 7d) suggest that the filtration of arsenic species by the membrane is an oxidation process.35

21

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 7. Results obtained by XPS analysis of the Cu/CNT membrane before and after arsenic filtration: (a) atomic percentages of Cu, C, N, O and As, (b) Cu2p spectra, (c) As3d spectra and (d) O1s spectra. The feed solution contained 100 µg L-1 of As(III) in the background solution of 0.01 M NaCl with an adjusted pH of 7.0.

3.7. As speciation results for the membrane permeate. Figure 8 displays As removal by the Cu/CNT membrane obtained at varying pH values. When the pH of the feed solution is controlled at 7.0, the concentration of As(III) in the feed solution is 176 µg L-1 while As(V) is found at a relatively low concentration of 20 µg L-1. After the filtration with the Cu/CNT membrane, As(V) is not detected in the permeate and the concentration of As(III) is consistently below 5 µg L-1. Overall, the removal efficiency of As(III) by the Cu/CNT membrane remains at above 97% at pH of 7.0 22

ACS Paragon Plus Environment

Page 22 of 46

Page 23 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

and all As(V) present in the feed solution are sequestrated. These results are comparable with those obtained with XPS analysis (Figure 7) as both experiments are performed at similar pH values and ionic strengths. Clearly, direct oxidation of As(III) to As(V) is limited in the feed solution, and therefore, should occur primarily during the filtration process. When the pH is adjusted to 6.0, arsenic in the feed solution exists solely as As(III) and at a mass concentration of 119.4 µg L-1. The acidic pH further inhibits oxidation of arsenite in the aqueous phase. As a result, no As(V) or organoarsenic is detected in the feed solution. After the feed solution is filtered by the Cu/CNT membrane, the concentration of As(III) in the permeate increases gradually with the increase in cumulative permeate volume, and the final removal efficiency decreases to 50% as the cumulative permeate volume reaches 40 mL or 81.6 L m-2 as unit permeate throughput. On the other hand, neither As(V) nor organoarsenic is detected in the permeate throughout the filtration experiment. In comparison, when the pH of the feed solution increases to an alkaline value of 8.0, As(III) concentration decreases to 78 µg L-1, while that of As(V) increases to 26.7 µg L-1, due to enhanced oxidation of As(III) by dissolved oxygen in water. In this case, As(III) is removed by Cu/CNT membrane to below 3 µg L-1, and As(V) is not detected in the permeates despite the relatively high concentration of As(V) in the feed solution. The results obtained at three pH levels demonstrate that the Cu/CNT membrane is very effective in removing As(III) and As(V) in neutral to alkaline pH conditions, but less effective for As(III) in acidic conditions. This finding is in agreement to that obtained with CuO nanoparticles. Martinson and Reddy reported that the As(III) removal efficiency obtained by CuO nanoparticles was the highest at the pH value of 9.3 and the lowest at pH of 6.1.36

23

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 8. As removal by the Cu/CNT membrane at varied pH values. When pH = 6.0, As(III) in feed solution = 119.4 µg L-1; when pH = 7.0, As(III) in feed solution = 176 µg L-1, As(V) in feed solution = 20.3 µg L-1; and, when pH = 8.0, As(III) in feed solution = 78 µg L-1, As(V) in feed solution = 26.7 µg L-1; flow rate = 0.05 mL min-1, ionic strength = 0.01 mol L-1. As(V) was not detected in any permeate samples.

3.8. CV curves of Cu-sputtered membranes. In this study, CV tests are carried out to determine potential oxidation of As(III) by the studied membranes. As shown in Figure 9a, the CV curve for the Cu/MCE membrane shows no distinctive peak, which meant that no oxidation reaction occurred between the arsenite solution and the membrane. On the contrary, a weak peak is observed on the CV curve at 0.2 V (vs. Ag/AgCl) for the CNT membrane (Figure 9b). Comparatively, distinct peak appears between 0.2 - 0.3 V (vs. Ag/AgCl) in the CV curve for the Cu/CNT membranes (Figure 9c). According to Aguirre et al,37 this CV peak corresponds to the oxidation of As(III) to As(V), indicating that the Cu/CNT membranes exhibits stronger oxidation of As(III) to As(V) than CNT membranes. The appearance of the CV peak at 0.2 - 0.3 V for the Cu/CNT membrane is 24

ACS Paragon Plus Environment

Page 24 of 46

Page 25 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

coincident with its high arsenic removal efficiencies (Figure 3). These results further confirm the occurrence of As(III) oxidation to As(V) during Cu/CNT membrane filtration of arsenite solutions. Moreover, the oxidation reaction is stronger for the Cu/CNT membrane than that for the Cu/MCE membrane, suggesting enhanced oxidation reaction due to higher electrical conductance of CNTs than that of MCE membranes.

25

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 9. Cyclic voltammetry curves for (a) the Cu/MCE membrane and (b) the CNT membrane and (c) Cu/CNT membrane measured in a solution containing 1.0 mg L-1 26

ACS Paragon Plus Environment

Page 26 of 46

Page 27 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

As(III) and 0.01 M NaCl at a scan rate of 2 mV s-1. 3.9. Arsenic filtration mechanism. On the basis of above analyses, it is reasonable to deduce that As(III) removal by the Cu/CNT membrane involves two major steps. During the first step, As(III) in the feed solution is oxidized into As(V) by dissolved oxygen under the catalysis of magnetron-sputtered Cu nanoparticles. In the second step, the formed As(V) species are adsorbed by the Cu/CNT membrane (Figure 10). The existence of As(III) on the membrane surface also indicates potential occurrence of direct adsorption of As(III) by the Cu/CNT membrane (Figure 7c) although its contribution to the total adsorption is minor as compared to As(V) adsorption. Overall, the magnetron-sputtered Cu nanoparticles play a critical role in arsenite removal. The catalytic effect of Cu nanoparticles possibly owes to the Cu(II)/Cu(I) couple with Cu atoms acting as electron mediation centers.38, 39 It is speculated that electron transfer in unconstrained Cu(II/I) systems includes the rupture of one or two coordinate bonds as well as a twisting of the remaining bond angles. Therefore, internal reorganizational energies are assumed to be significant in the electron-transfer behavior of copper.39 Furthermore, the fast electron transfer kinetics on multi-walled carbon nanotube could facilitate the oxidation of As(III).40 As such, arsenic oxidation by molecular oxygen is promoted and the reaction time is reduced from several days to the short timeframe of the filtration (< 1 second). In addition to the catalytic effect, cupric hydroxides formed through oxidation of Cu(0) further serves as the adsorbent for the formed arsenate. Similarly, Bang et al. reported that the removal of As(III) by Fe(0) resulted from the adsorption of As(III) on ferric hydroxides formed through oxidation of Fe(0) by dissolved oxygen.41 Furthermore, the more As(III) oxidized to As(V), the higher the As removal rate, owing to the fact that As(V) is less soluble than As(III). Therefore, the Cu/CNT membranes remove more arsenite at neutral to alkaline pHs than at the acidic pH 27

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(Figure 8). Also, the Cu/CNT membranes are most efficient in oxidizing As(III) to As(V), thus having the highest As removal efficiency among the membranes evaluated in this study. Overall, this study indicates that the Cu/CNT composite on the membrane plays a dual role of catalyst and adsorbent in the remediation of As-contaminated water.

Figure 10. Schematic diagram of arsenic filtration mechanism by the Cu/CNT membrane.

3.10. Applicability of magnetron sputtering to membrane water treatment. The aforementioned success of the Cu/CNT membrane for arsenic removal demonstrates that magnetron sputtering is potentially a viable technique for developing metal-CNT membranes tailored for water treatment. Indeed, various metal targets, such as Cu, Ag, Al and some metal oxides like MoS2 and ZnO, have been used for sputtering other substrate surfaces.13, 15, 42, 43 Similar approaches have been widely used in the modification of organic or biological materials with metallic nanomaterials as noted above. In addition to the versatility in coating materials, magnetron sputtering is easier to operate than chemical functionalization and can be used for flat-sheet membranes in large-scale applications. In this study, 10 minutes of sputtering appeared to be sufficient in producing a stable thin film of copper 28

ACS Paragon Plus Environment

Page 28 of 46

Page 29 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

nanoparticles on the surfaces of the MCE membranes and the CNT-modified membranes, without any noticeable damages on the substrate membranes (Figure 1). The coated Cu layer exhibits promising mechanical stability (Figure S1) and good resistance to corrosive dissolution of Cu (Figure S4). Currently, there is an urgent need to develop novel materials for affordable, energy-efficient, ease of operation, and safe water purification for human beings.1 To this end, the Cu/CNT membrane prepared by the magnetron sputtering technique have demonstrated some merits, such as (1) improved energy efficiency and high water permeability for arsenic filtration as they can be operated at applied pressures of 0.01 bar (Figure S5), which is much lower than existing RO membranes;44 (2) minimal environmental impact during production as the active Cu component is facilely deposited onto the substrate membrane without the use of any harmful chemicals; (3) simple in the production process; and, (4) effective for targeted contaminants as described above. These merits truly open the door for scaled-up production of functional CNT membranes for water treatment purposes.

4. CONCLUSIONS This study demonstrates that magnetron-sputtering technique is potentially a viable approach for the development of functional Cu-CNT membranes for drinking water treatment. By sputtering Cu onto a CNT membrane, stable and highly permeable Cu/CNT composite is formed in a facile way. Compared to the substrate MCE membrane, the polymeric membrane sputtered with Cu, or the polymeric membrane coated with CNT, the Cu/CNT membrane shows the best capacity in eliminating arsenite from aqueous solutions under varying pH and ionic strength conditions. However, competing anions (sulfate and phosphate) play negative roles in the removal of As(III), and therefore, pretreatment of the source water may be needed 29

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

for the treatment of source waters containing competing anions. Based upon XPS and HPLC-ICP-MS results, As removal by the Cu/CNT membrane primarily involves two major steps: (1) As(III) oxidation to As(V) by aqueous oxygen under the catalysis of sputtered Cu nanoparticles, and (2) the adsorption of As(V) onto copper hydroxides on the composite membrane.

ASSOCIATED CONTENTS Supporting Information. Digital photos showing the flexibility of CNT and Cu/CNT membranes. (a - b) a CNT membrane; (c - d) a Cu/CNT membrane, SEM-EDS spectra of Cu nanoparticles on the surface of Cu/CNT membranes with sputtering time of (a) 5, (b) 10 and (c) 15 min, SEM image of the Cu/CNT membrane after treatment with As(III), Concentrations of Cu ions released in the permeate by various Cu-sputtered membranes, Water flux as a function of transmembrane pressure for different membranes. AUTHOR INFORMATION Corresponding Author * Email: [email protected] Phone: +86 10 5880 7734. Fax: +86 10 5880 7734. Notes The authors declare no competing financial interest. ACKNOWLEDGMENT The author would like to thank Prof. Zhiqiang Tan for providing access to the HPLC-ICP-MS system and Mr. Jingwen Xu for performing arsenic speciation analysis with the system. This work was financially supported by the Fundamental Research Funds for the Central Universities of China (310421111).

REFERENCES 30

ACS Paragon Plus Environment

Page 30 of 46

Page 31 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

[1] Lee, A.; Elam, J. W.; Darling, S. B. Membrane materials for water purification: design, development, and application. Environ. Sci.: Water Res. Technol. 2016, 2, 17-42. [2] Huang, H.; Schwab, K.; Jacangelo, J. G. Pretreatment for Low Pressure Membranes in Water Treatment: A Review. Environ. Sci. Technol. 2009, 43, 3011-3019. [3] Shanmuganathan, S.; Loganathan, P.; Kazner, C.; Johir, M. A. H.; Vigneswaran, S. Submerged membrane filtration adsorption hybrid system for the removal of organic micropollutants from a water reclamation plant reverse osmosis concentrate. Desalination 2017, 401, 134-141. [4] Shanmuganathan, S.; Johir, M. A. H.; Nguyen, T. V.; Kandasamy, J.; Vigneswaran, S. Experimental evaluation of microfiltration–granular activated carbon (MF– GAC)/nano filter hybrid system in high quality water reuse. Journal of Membrane Science 2015, 476, 1-9. [5] Das, R.; Eaqub Ali, Md.; Sharifah Bee Abd Hamid.; Ramakrishna, S.; Chowdhury, Z. Z. Carbon nanotube membranes for water purification: A bright future in water desalination. Desalination 2014, 336, 97-109. [6] Ma, P.C.; Siddiqui, N. A.; Marom, G.; Kim, J. K. Dispersion and functionalization of carbon nanotubes for polymer-based nanocomposites: A review. Composites: Part A 2010, 41, 1345-1367. [7] Hooijdonk, E. V.; Bittencourt, C.; Snyders, R.; Colomer, J. F. Functionalization of vertically aligned carbon nanotubes. Beilstein J. Nanotechnol. 2013, 4, 129-152. [8] Bakshi, S. R.; Lahiri, D.; Agarwal, A. Carbon nanotube reinforced metal matrix composites - a review. International Materials Reviews 2010, 55, 41-64. [9] Hu, X. G.; Dong, S. J. Metal nanomaterials and carbon nanotubes—synthesis, 31

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

functionalization and potential applications towards electrochemistry. J. Mater. Chem. 2008, 18, 1279-1295. [10] Rossnagel, S. M. In Opportunities for Innovation: Advanced Surface Engineering; Sproul, W. D., Legg, K. O., Eds.; Technomic Publishing Co: Switzerland, 1995; Chapter 2, pp 7-28. [11] Yip, J.; Jiang, S. Q.; Wong, C. Characterization of metallic textiles deposited by magnetron sputtering and traditional metallic treatments. Surface & Coatings Technology 2009, 204, 380-385. [12] Kelly, P. J.; Arnell, R. D. Magnetron sputtering: a review of recent developments and applications. Vacuum 2000, 56, 159-172. [13] Lv, P. F.; Wei, A. F.; Wang, Y. W.; Li, D. W.; Zhang, J.; Lucia, L. A.; Wei, Q. F. Copper nanoparticles-sputtered bacterial cellulose nanocomposites displaying enhanced electromagnetic shielding, thermal, conduction, and mechanical properties. Cellulose 2016, 23, 3117-3127. [14] Pham, K. C.; McPhail, D. S.; Mattevi, C.; Wee, A. T. S.; Chua, D. H. C. Graphene-Carbon Nanotube Hybrids as Robust Catalyst Supports in Proton Exchange Membrane Fuel Cells. Journal of The Electrochemical Society 2016, 163, F255-F263. [15] Jiang, S. X.; Qin, W. F.; Guo, R. H.; Zhang, L. Surface functionalization of nanostructured silver-coated polyester fabric by magnetron sputtering. Surface & Coatings Technology 2010, 204, 3662–3667. [16] Matatov-Meytal, Yurii I.; Sheintuch, Moshe. Catalytic Abatement of Water Pollutants. Ind. Eng. Chem. Res. 1998, 37, 309-326. [17] Wu, Z. C.; Chen, Z. H.; Du, X.; Logan, J. M.; Sippel, J.; Nikolou, M.; Kamaras, K.; Reynolds, J. R.; Tanner, D. B.; Hebard, A. F.; Rinzler, A. G. Transparent, 32

ACS Paragon Plus Environment

Page 32 of 46

Page 33 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

conductive carbon nanotube films. Science 2004, 305, 1273-1276. [18] Espinós, J. P.; Morales, J.; Barranco, A.; Caballero, A.; Holgado, J. P.; González-Elipe, A. R. Interface Effects for Cu, CuO, and Cu2O Deposited on SiO2 and ZrO2. XPS Determination of the Valence State of Copper in Cu/SiO2 and Cu/ZrO2 Catalysts. J. Phys. Chem. B 2002, 106, 6921-6929. [19] Tahir, D.; Tougaard, S. Electronic and optical properties of Cu, CuO and Cu2O studied by electron spectroscopy. J. Phys.: Condens. Matter 2012, 24, 1-8. [20] Payne, K. B.; Abdel-Fattah, T. M. Adsorption of Arsenate and Arsenite by Iron-Treated Activated Carbon and Zeolites: Effects of pH, Temperature, and Ionic Strength. Journal of Environmental Science and Health 2005, 40, 723-749. [21] Chen, J.; Wang, J. Y.; Zhang, G. S.; Wu, Q. Y.; Wang, D. T. Facile fabrication of nanostructured cerium-manganese binary oxide for enhanced arsenite removal from water. Chem. Eng. J. 2018, 334, 1518-1526. [22] Lee, H.; Kim, D.; Kim, J.; Ji, M. K.; Han, Y. S.; Park, Y. T.; Yun, H. S.; Choi, J. As(III) and As(V) removal from the aqueous phase via adsorption onto acid mine drainage sludge (AMDS) alginate beads and goethite alginate beads. Journal of Hazardous Materials 2015, 292, 146-154. [23] Escudero, C.; Fiol, N.; Villaescusa, Isabel.; Bollinger, J. C. Arsenic removal by a waste metal (hydr)oxide entrapped into calcium alginate beads. Journal of Hazardous Materials 2009, 164, 533-541. [24] Cao, A. M.; Monnell, J. D.; Matranga, C.; Wu, J. M.; Cao, L. L.; Gao, Di. Hierarchical Nanostructured Copper Oxide and Its Application in Arsenic Removal. J. Phys. Chem. C 2007, 111, 18624-18628. [25] Liu, H.; Zuo, K. C.; Vecitis, C. D. Titanium Dioxide-Coated Carbon Nanotube

Network Filter for Rapid and Effective Arsenic Sorption. Environ. Sci. Technol. 33

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2014, 48, 13871-13879. [26] Hlavay, J.; Polya´k, K. Determination of surface properties of iron hydroxide-coated alumina adsorbent prepared for removal of arsenic from drinking water. Journal of Colloid and Interface Science 2005, 284, 71-77. [27] Goswami, A.; Raul, P. K.; Purkait, M. K. Arsenic adsorption using copper (II) oxide nanoparticles. Chemical Engineering Research and Design 2012, 90, 1387-1396. [28] Pillewan, P.; Mukherjee, S.; Roychowdhury, T.; Das, S.; Bansiwal, A.; Rayalu, S. Removal of As (III) and As (V) from water by copper oxide incorporated mesoporous alumina. Journal of Hazardous Materials 2011, 186, 367-375. [29] Pillewan, P.; Mukherjee, S.; Meher, A. K.; Rayalu, S.; Bansiwal, A. Removal of Arsenic (III) and Arsenic (V) Using Copper Exchange Zeolite-A. Environmental Progress & Sustainable Energy 2014, 33, 1274-1282. [30] Kanel, S. R.; Manning, B.; Charlet, L.; Choi, H. Removal of Arsenic(III) from Groundwater by Nanoscale Zero-Valent iron. Environ. Sci. Technol. 2005, 39, 1291-1298. [31] Zhang, G.; Qu, J.; Liu, H.; Liu, R.; Li, G. Removal mechanism of As(III) by a novel Fe-Mn binary oxide adsorbent: oxidation and sorption. Environ. Sci. Technol. 2007, 41, 4613-4619. [32] Zhang, L.; Zhu, T.; Liu, X.; Zhang, W. Simultaneous oxidation and adsorption of As(III) from water by cerium modified chitosan ultrafine nanobiosorbent. J. Hazard. Mater. 2016, 308, 1-10. [33] Yu, Y.; Zhang, C.; Yang, L.; Paul Chen, J. Cerium oxide modified activated carbon as an efficient and effective adsorbent for rapid uptake of arsenate and arsenite: material development and study of performance and mechanisms. Chem. 34

ACS Paragon Plus Environment

Page 34 of 46

Page 35 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Eng. J. 2017, 315, 630-638. [34] Henke, K. Arsenic: Environmental Chemistry, Health Threats, and Waste Treatment, 1st ed; Wiley: West Sussex, 2009. [35] Ge, X.; Ma, Y.; Song, X. Y.; Wang, G. Z.; Zhang, H. M.; Zhang, Y. X.; Zhao, H. J. β-FeOOH Nanorods/Carbon Foam-Based Hierarchically Porous Monolith for Highly Effective Arsenic Removal. ACS Appl. Mater. Interfaces 2017, 9, 13480-13490. [36] Martinson, C.A.; Reddy, K.J. Adsorption of arsenic(III) and arsenic(V) by cupric oxide nanoparticles. J. Colloid Interface Sci. 2009, 336, 406-411. [37] Aguirre, María del C.; Rivas, B. L.; Basáez, L.; Carlos, Peña-Farfal. Electrochemical Detection of Arsenite with Silver Electrodes in Inorganic Electrolyte and Natural System Mixtures. J. Braz. Chem. Soc. 2011, 22, 2362-2370. [38] Gawande, M. B.; Goswami, A.; Felpin, F. X.; Asefa, T.; Huang, X. X.; Silva, R.; Zou, X. X.; Zboril, R.; Varma, R. S. Cu and Cu-Based Nanoparticles: Synthesis and Applications in Catalysis. Chem. Rev. 2016, 116, 3722-3811. [39] Rorabacher, D. B. Electron Transfer by Copper Centers. Chem. Rev. 2004, 104, 651-698. [40] Nugent, J. M.; Santhanam, K. S. V.; Rubio, A.; Ajayan, P. M. Fast Electron Transfer Kinetics on Multiwalled Carbon Nanotube Microbundle Electrodes. Nano Letters 2001, 1, 87-91. [41] Bang, S.; Johnson, M. D.; Korfiatis, G. P.; Meng, X. G. Chemical reactions between arsenic and zero-valent iron in water. Water Res. 2005, 39, 763-770. [42] Suchea, M.; Christoulakis, S.; Katsarakis, N.; Kitsopoulos, T.; Kiriakidis, G. 35

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Comparative study of zinc oxide and aluminum doped zinc oxide transparent thin films grown by direct current magnetron sputtering. Thin Solid Films 2007, 515, 6562-6566. [43] Li, H.; Xie, M. L.; Zhang, G. G.; Fan, X. Q.; Li, X.; Zhu, M. H.; Wang, L. P. Structure and tribological behavior of Pb-Ti/MoS2 nanoscaled multilayer films deposited by magnetron sputtering method. Applied Surface Science 2018, 435, 48-54. [44] Misdan, N.; Lau, W. J.; Ismail, A. F. Seawater reverse osmosis (SWRO) desalination by thin-film composite membrane—current development, challenges and future prospects. Desalination 2012, 287, 228-237.

Table of Contents Graphic

36

ACS Paragon Plus Environment

Page 36 of 46

Page 37 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

37

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1. Scanning electron micrograms of the top surfaces of (a) the MCE membrane with high resolution (×50,000), scale bar = 500 nm; (b) the Cu/MCE membrane; (c) the

ACS Paragon Plus Environment

Page 38 of 46

Page 39 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

CNT membrane, with low resolution (×5,000), scale bar = 5 μm; (d) the Cu/CNT membrane with low resolution (×5,000); (e) the CNT membrane with high resolution (×50,000), scale bar = 500 nm; (f) the CNT membrane sputtered with Cu for 10 min with high resolution (×50,000); (g) the CNT membrane sputtered with Cu for 5 min with high resolution (×50,000); (h) the CNT membrane sputtered with Cu for 15 min with high resolution (×50,000).

Figure 2. XPS spectra of (a) the Cu/MCE membrane and (b) the Cu/CNT membrane, and high-resolution Cu2p spectra of (c) the Cu/MCE membrane and (d) the Cu/CNT membrane.

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3. Removals of arsenic spiked in the ultrapure water by membranes without Cusputtering (a), and spiked in background solutions with different NaCl concentrations by Cu-sputtered membranes (b); effects of sputtering time on As(III) removal with Cu/CNT membranes with background solution of 0.01 M NaCl (c). NaCl concentrations = 0, 0.01 and 0.1 mol L-1, initial As(III) concentration = 100 μg L-1, pH = 7.0, temperature = 25  1 °C.

Figure 4. Sorption kinetics of the Cu/CNT membrane for the removal of As(III). (a)

ACS Paragon Plus Environment

Page 40 of 46

Page 41 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

pseudo-first-order kinetic model; (b) pseudo-second-order kinetic model. Experimental conditions: arsenic concentration = 4.74 ppm, pH = 7.0, constant permeate flux = 6.7 L m-2 h-1, and temperature 25 °C.

Figure 5. Effects of sulfate and phosphate on the removal of As(III). Initial As concentration = 100 ppb, pH = 7.0, temperature 25 °C.

Figure 6. As(III) removal with Cu/CNT membrane for different filtration-regeneration cycles. Initial As(III) concentration = 100 ppb; regenerated by 0.001 M NaOH; constant permeate flux = 6.7 L m-2 h-1; permeate volume of each sample = 40 mL.

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 7. Results obtained by XPS analysis of the Cu/CNT membrane before and after arsenic filtration: (a) atomic percentages of Cu, C, N, O and As, (b) Cu2p spectra, (c) As3d spectra and (d) O1s spectra. The feed solution contained 100 μg L-1 of As(III) in the background solution of 0.01 M NaCl with an adjusted pH of 7.0.

Figure 8. As removal by the Cu/CNT membrane at varied pH values. When pH = 6.0,

ACS Paragon Plus Environment

Page 42 of 46

Page 43 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

As(III) in feed solution = 119.4 μg L-1; when pH = 7.0, As(III) in feed solution = 176 μg L-1, As(V) in feed solution = 20.3 μg L-1; and, when pH = 8.0, As(III) in feed solution = 78 μg L-1, As(V) in feed solution = 26.7 μg L-1; flow rate = 0.05 mL min-1, ionic strength = 0.01 mol L-1. As(V) was not detected in any permeate samples.

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 9. Cyclic voltammetry curves for (a) the Cu/MCE membrane and (b) the CNT membrane and (c) Cu/CNT membrane measured in a solution containing 1.0 mg L-1 As(III) and 0.01 M NaCl at a scan rate of 2 mV s-1.

ACS Paragon Plus Environment

Page 44 of 46

Page 45 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 10. Schematic diagram of arsenic filtration mechanism by the Cu/CNT membrane.

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Graphic abstract

ACS Paragon Plus Environment

Page 46 of 46