Tuned Fabrication of the Aligned and Opened CNT ... - ACS Publications

Subscriber access provided by University of South Dakota

Communication

Tuned Fabrication of the Aligned and Opened CNT Membrane with Exceptionally High Permeability and Selectivity for Bioalcohol Recovery Decai Yang, Dong-Xu Tian, Chuang Xue, Fei Gao, Yang Liu, Hong Li, Yongming Bao, Jingjing Liang, Zongbin Zhao, and Jieshan Qiu Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.8b01831 • Publication Date (Web): 22 Aug 2018 Downloaded from http://pubs.acs.org on August 23, 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 24 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

Nano Letters

Tuned Fabrication of the Aligned and Opened CNT Membrane with Exceptionally High Permeability and Selectivity for Bioalcohol Recovery Decai Yang,†,║ Dongxu Tian,‡,║ Chuang Xue,*,† ,║ Fei Gao,¶ Yang Liu,§ Hong Li,¶ Yongming Bao,† Jingjing Liang,§ Zongbin Zhao,§ and Jieshan Qiu§ †

School of Life Science and Biotechnology, Dalian University of Technology, Dalian 116024,

China ‡

School of Chemistry, State Key Laboratory of Fine Chemicals, Dalian University of

Technology, Dalian 116024, China ¶

Key Laboratory of Materials Modification by Laser, Ion, and Electron Beams (Ministry of

Education), School of Physics, Dalian University of Technology, Dalian, 116024, China §

State Key Laboratory of Fine Chemicals, Liaoning Key Laboratory for Energy Materials and

Chemical Engineering, PSU-DUT Joint Center for Energy Research, School of Chemical Engineering, Dalian University of Technology, Dalian 116024, China

*Corresponding author: [email protected]

ACS Paragon Plus Environment

1

Nano Letters 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

Page 2 of 24

ABSTRACT: Synthetic membranes usually suffer from a ubiquitous tradeoff between permeability and selectivity. Carbon nanotube (CNT)-based hybrid materials have shown attractive properties in high-performance membrane preparation; however, the aggregation of random CNTs in polymer remains a great challenge. Herein, the aligned and open-ended CNT/(polydimethylsiloxane) PDMS membranes are controllably fabricated to form a hamburger-like structure that possesses nanochannels (~10 nm) in the intermediate layer as well as angstrom cavities in the embedded PDMS. These aligned CNT membranes surpass the filling content limitation of the nonaligned CNT/PDMS membrane (37.4 wt % vs. ~10 wt %), leading to excellent mechanical properties and a multiplying performance increase of mass flux and selectivity for the separation of alcohols. The membranes break the permeability-selectivity tradeoff, with both parameters remarkably increasing (maximum 9 times) for bioalcohol separation. The established pervaporative-ultrafiltration mechanism indicates that the penetrant molecules preferentially pass through CNT internal nanochannels with increasing membrane permeability, thereby paving a way to nanoscale design of highly efficient channeled membranes for separation application.

KEYWORDS: carbon nanotubes, composite membrane, biofuel, separation, pervaporation


2

Page 3 of 24 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

Nano Letters

Butanol and ethanol, as representative biofuels, are receiving enthusiastic attention because of the concerns regarding climate change and crude oil depletion.1-3 Furthermore, butanol is considered to be superior to ethanol, with advanced properties including higher energy density and better compatibility with petroleum, but its production is limited by low final butanol concentration in fermentation broth that results in high energy consumption for butanol recovery.4-6 The pervaporation is considered to be the most promising separation technology due to its energy efficiency; however, current pervaporative membranes suffer from a tradeoff between permeability and selectivity.7,8 The pervaporation (PV) membranes prepared with polydimethylsiloxane (PDMS) as the base material have a variety of potential uses in the separation of volatile organic chemicals (VOCs), water and gas.9,10 According to the solution-diffusion model, the diffusion process is the rate-limiting step in PV membranes, the rate of which is much lower than the rates of dissolution into the upstream and desorption from the downstream.11,12 Moreover, molecular dynamics (MD) simulations revealed two types of the penetrant molecules motions inside the PDMS polymers: a molecule “jumps” from one cavity in a polymer matrix to another and undergoes “oscillating motions” inside cavities.13,14 The lengths of the jumps and oscillating motions are only ~8-15 Å and ≤ 5 Å (depending on operational conditions), respectively, which are of the order of the penetrant molecule size, resulting in low diffusion coefficient and permeability. Hence, determining how to overcome the bottleneck of low diffusion during pervaporation is the key for improving the membrane separation performance. Carbon nanotubes (CNTs) are excellent materials with outstanding mechanical properties of a large specific surface area and excellent thermal stability.15-17 CNTs with a one-dimensional graphitic nanochannel structure are of particular interest as building blocks for new generation


3

Nano Letters 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

Page 4 of 24

membranes.18 Majumder et al. verified that the transport of liquids (water and alkanes) through the aligned CNTs membrane was orders of magnitude faster than conventional hydrodynamic flow.19 Therefore, the introduction of CNTs in polymer matrices provides a new orientation for composite membranes fabrication. Despite these high expectations, the CNT/PDMS mixed membrane fabricated by filling CNTs usually leads to a random aggregation and low filling content.20,21 Moreover, because of their small size, CNTs are normally curled and twisted, thereby limiting their contribution in changing the membrane structure. Although a number of excellent advancements in promoting the alignment of CNTs through surface-lattice-guided and electrical-field-assisted growth CNT sheets, etc., have been made recently, it remains a great challenge to efficiently control the uniform distribution of CNTs in the polymer matrix membrane. 22-24 The ex situ alignment of CNTs by their growth before composites fabrication offers a feasible route for controllably assembling organized nanocomposites.25-31 Hinds et al. first fabricated vertically aligned CNT (VACNT) membranes using CNTs with an inner core diameter of ca. 6 nm embedded in a porous-free polystyrene (PS) matrix.26 The membranes in which 1.6nm-wide double-walled CNT arrays serve as only through pores were fabricated to span a silicon nitride (Si3N4) matrix.27 In summary, VACNT membranes have usually been embedded or spanned with an impermeable matrix (e.g., parylene-N, PS, Si3N4, epoxy and polyurethane), in which CNTs with pore sizes of several nanometers were the only channels to allow water transport and block [Ru(NH3)6]3+, [Fe(CN)6]3-, dye, Au nanoparticles, virus and bacteria with nano- or microsizes (Table S1).28-31 All the abovementioned aligned CNT/impermeable polymer composite membranes are excellent candidates for use in ultrafiltration or nanofiltration but are not qualified for various VOCs separations because of their nanometer-scale pore sizes.


4

Page 5 of 24 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

Nano Letters

Here, for the first time, we present a tuned method for fabricating vertically aligned and open-ended CNT/PDMS composite membranes with both surface sides coated with an ultrathin PDMS layer. The membrane has a unique hamburger-like structure that possesses nanochannels (~10 nm) in the intermediate layer for use in ultrafiltration as well as angstrom-cavities for use in pervaporation, which essentially surpass the diffusion limitation underlying an integrative pervaporation-ultrafiltration mechanism. The penetrant molecules can permeate through the angstrom-scale cavities of the PDMS layer on the top of the membrane via the pervaporation mechanism and then preferentially flow into the CNT nanochannels at a faster rate via the ultrafiltration mechanism. Figure 1a illustrates the fabrication steps of the aligned CNT (open-ended)/PDMS composite membrane; the details are provided in Figure S1. Both multiwalled and single-walled CNTs could be utilized to fabricate the composite membranes. Here, we demonstrate multiwalled CNTs due to their ease of synthesis. The CNTs are highly aligned (Figure 1b), with the range of the inner diameters of the CNTs of 3-17 nm (9.67 nm average; Figure 2a), which is an order of magnitude larger than typical molecular diameters, such as water, n-butanol, and ethanol but smaller than typical dimensions of bacteria and viruses (Table S2). Thus, our CNT composite membrane could block biological threats and facilitate molecular particles permeation. Since some of the CNTs may have typical defects, such as kinks, metallic impurities and “bamboo-type” inner structures, the annealing treatment was used to reduce the metal catalyst nanoparticles and the structural defects.32,33 As shown in Figure S2, the annealing treatment can efficiently remove metal particles and lower defects in the hollow core or at the tips of the CNTs. The annealed CNTs form a better hollow tubular structure than the untreated CNTs (Figure S2, S3a). Furthermore, the IG/ID ratio of CNTs can be significantly increased from


5

Nano Letters 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

Page 6 of 24

2.27 to 4.15 by the high temperature annealing treatment (Figure S2d,h); thus, the annealed CNTs possess significantly lower defects than the untreated CNTs. The Ar plasma etching process under our optimized condition effectively opened the CNT ends without any negative effect on the walls of the CNTs and their alignment (Figure 1c, Figure S3b and S4a). Therefore, the internal pipeline of CNTs can be the ideal channel for transportation of molecules. Due to the agglomeration of the top ends of the CNTs induced by increasing Ar plasma treatment time, we selected an etching time of 10 min to effectively etch the disordered carbons and open the top ends of the as-grown CNTs (Figure S5).34,35 The PDMS was well-filled into the gaps of the CNTs via spin-coating, forming a defect-free composite membrane structure (Figure S4b) without breaking the aligned structure of the CNTs (Figure 1d). Since PDMS on the substrate side (Figure S4d) must be etched in advance to guarantee the CNT ends are open, the combined Ar/CF4 plasma was employed to remove a thin layer of excess polymer (Figure 1e) and then open the CNT ends.36 More importantly, the optimized etching time (20 min) was chosen to ensure that the CNT ends were open while maintaining satisfactory mechanical properties of the composite membrane (Figure S6; Table S3). However, after Ar/CF4 mixture plasma treatment, the water contact angle and separation performance of the membranes reduced because of the roughened surface of the membrane (Figure S4e; Table S4, S5). Therefore, to compensate for the hydrophobicity loss (Table S4), an additional thin layer of PDMS (ca. 10 µm) was applied using the spin-coating technique to restore both the water contact angle and the separation performance by smoothing the membrane surface (Figure 1f and Figure S4f). In addition, the thickness of the composite membrane could be controlled from micrometers to millimeters by regulating the height of the CNT arrays or sliced into transparent thin films with nanometers using an ultramicrotome.


6

Page 7 of 24 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

Nano Letters

Because the formation and stabilization of the CNT-bonded macroradicals can contribute to the increase of the oxidative thermal degradation temperature of polymer,37 the aligned CNT arrays can improve the thermal stability of PDMS membrane (Figure S7). The disordered/random CNTs usually have a much lower filling content in PDMS because of their inevitably severe agglomeration, and filling more random CNTs (> 10 wt %) into PDMS can cause membrane defect or leakage, which significantly hinders the separation performance of composite membrane.21 In the present study, the proportion of CNTs loading in the PDMS composite membrane is calculated as 37.36 wt % and the estimated areal density is 2.23 × 1011 CNTs per cm2, which are three-fold higher than maximum loading of disordered CNTs in PDMS. Although CNTs functionalization can improve the dispersion and solubility behavior of CNTs,38,39 this chemical modification might introduce CNTs defects, thus undermining the membrane separation performance. For example, the acidic oxidation treatment can introduce functional groups (e.g., carboxyl groups) onto the CNTs surface; however, this method creates many more defective sites on the sidewalls of CNTs due to acids cutting the CNTs into short pieces.40,41 Adopting the CNT arrays significantly improves the compatibility with PDMS, thus breaking through the maximum mass fraction limitation of the filling of the disordered CNTs. In addition, the content of CNTs in the PDMS polymer matrix can be adjusted or enhanced by varying the CNT array densities through tuning the growth parameters. Introduction of aligned CNTs can remarkably reinforce the mechanical properties of PDMS, which has good vibration/motility in the polymer chain but relatively poor mechanical properties (Figure 2b). The enhancements are predominantly attributed to the favorable interfacial structure between PDMS and aligned CNT arrays and can be favorable for stress transfer across the CNTs/PDMS interface with more efficient load transfer.22 The aligned CNT arrays in the PDMS membrane


7

Nano Letters 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

Page 8 of 24

could provide a self-supporting structure that is superior to the traditional composite membranes with various coated supporting layers.42-44 Therefore, the high loading of CNTs and the outstanding mechanical property and thermal stability of the aligned CNT (open-ended)/PDMS membrane are beneficial for various chemicals separations. We representatively demonstrate the unexpectedly high performance of our aligned CNTs/PDMS composite membranes for butanol and ethanol separations (Figure 3a,b). The aligned CNT (closed-ended)/PDMS membrane in the PV process has higher total flux, butanol flux and separation factor with an enhancement of 51.8 %, 74.8 % and 20.4 %, respectively, compared to those of the nonaligned CNT/PDMS membrane. In particular, the aligned CNT (open-ends)/PDMS membrane after annealing breaks the permeability-selectivity tradeoff, allowing the total flux, butanol flux and butanol separation factor to simultaneously increase by 494.8 %, 627.0 % and 30.4 %, respectively, compared to the pure PDMS membrane. The vapor stripping-vapor permeation (VSVP) process in which molecule vapors are at both sides of the membrane has 755.5 % and 914.0 % increase in total flux and butanol flux, respectively, with the butanol separation factor (105.0) increasing by 49.5 % for the overall VSVP process. For the VSVP process, molecules in vapor form at the feed side of the membrane more easily dissolve into/pass through the membrane and have a high diffusion rate during VSVP process.45 Therefore, the mass flux in the VSVP process is higher than that in the PV process. Since metallic impurities and structural defects in the hollow core or at the tips of the CNTs may block the penetrant molecules through the CNT channels (Figure S2), the annealing method was used to improve the quality of the CNTs.32,33 As expected, the performance of the aligned CNT (open-ended)/PDMS membrane being annealed is 1.5 times higher than that of the unannealed membrane, which is 3.7 times higher than that of the “closed CNT” membrane in the


8

Page 9 of 24 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

Nano Letters

VSVP process, respectively (Figure S8). A similar performance improvement was also observed for bioethanol recovery (Figure S8), indicating that the annealing treatment could make CNT channels more permeable for alcohols and water by reducing metal catalyst nanoparticles and structural defects in the CNT channels. These results suggest that, when the membrane is exposed to a 1.5 wt % feed solution of n-butanol from a conventional fermentation, approximately 53.8 wt % of butanol can be recovered in the permeate. The membrane in this study exhibited the best butanol separation performance in the VSVP process when compared with other reported membranes with the same level of membrane thickness (Figure 3c; Table S6), thereby saving capital investment and reducing the footprint for constructing a membrane bioreactor. Furthermore, the mass flux of the CNT membrane can be easily enhanced by decreasing the membrane thickness to several micrometers. The hydrophobic CNTs filled into PDMS membrane could enhance both the water contact angle and the hydrophobicity of the membrane (Table S4). However, the random or selfaggregating CNTs usually provide an irregular alternative route for mass transport that contributes to the limited improvement of the membrane performance (Figure S9).38 The significant improvements in the aligned CNT (closed-ended) membrane were attributed to its excellent hydrophobic property with the higher contact angle of 132° (Table S4). Moreover, the membrane exhibits a higher flux, which can be attributed to the larger filling mass fraction and more orderly distribution of CNTs, allowing molecules transport along the walls of the CNTs (Figure 4b). Once both ends of the CNTs are opened, the aligned CNT/PDMS membrane would form internal nanochannels (3-17 nm diameters), such as the conventional ultrafiltration membrane coated with an ultrathin PDMS layer. To confirm whether the penetrant molecules permeate through the inner CNT channels, both sides of the aligned CNT (open-ended)/PDMS


9

Nano Letters 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

Page 10 of 24

membrane were etched by Ar/CF4 to expose the CNTs from the PDMS polymer (Figure S10). As shown in Table S1, the water permeability of the aligned CNT (open-ended)/PDMS membrane after being etched at both sides was 853.5 (± 31.7) LMH/bar (equivalent to 8.53×105 g m-2 h-1) at 20 °C, which was three orders of magnitude higher than that of the unetched membrane (456.8 g m-2 h-1 in butanol/water solution). The penetrant molecules were found to freely pass through the inner nanochannels of the CNTs because the etched membrane possessed a high water permeability at the same level as those of other vertically aligned CNT composite membranes for ultrafiltration (Table S1). Based on the three-step solution-diffusion mechanism,11 the rate-limiting step is the molecule diffusion through the membrane and the local segmental dynamics of the open transient gaps (i.e., free-volume elements of polymer chains) through which small molecule diffusion occurs.8,12 In the PDMS matrix, the transient gaps or cavities for penetrant molecules are usually several angstroms in size, thereby severely restricting the diffusion of molecules (Figure 4a). The diffusion coefficients decrease with the increased size of penetrant molecules.46 The volumes of the CNT channels are an order of magnitude higher than the volumes of the PDMS cavities. As shown in Figure 4c, the open-ended CNTs mainly provide orderly nanochannels with average diameter of 9.67 nm, which easily facilitates the penetrant molecules (e.g., water and butanol with several angstroms) to diffuse through the membrane. Furthermore, the frictionless surfaces at the CNT walls could lead to high fluid velocities of molecules. In addition, based on the “gatekeeper” mechanism for controlling the selectivity of chemicals transport through the CNT membrane, functional groups can be introduced to the ends of the CNTs to make the CNTs more hydrophobic or hydrophilic,47,48 further improving the separation performance of the membrane. The aligned CNT (open-ended)/PDMS composite membrane also


10

Page 11 of 24 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

Nano Letters

exhibits an exponential increase (maximum 7 times) of mass flux and separation factor for bioethanol separation (Figure 3a,b). Therefore, the aligned CNT (open-ended)/PDMS membrane in the PV or VSVP process can achieve outstanding performance if integrated with the fermentation of biofuels (butanol or ethanol).49 The density functional theory (DFT) calculations in this study indicate that the penetrant molecules preferentially pass through the CNT internal channels, rather than the PDMS chains and the CNT external wall. As seen in Figure 5, the order of adsorption energy of molecules onto the PDMS chain and CNT is as follows: C4H9OH > C2H5OH > H2O. The PDMS and CNTs preferentially adsorb butanol rather than ethanol molecules in aqueous solution, which is in reasonable agreement with the above experimental results and the theoretical prediction based on the Hansen solubility parameters (Table S2). In particular, the adsorption energy of H2O onto the PDMS chain is greater than the adsorption energy onto the external and internal walls of the CNTs, indicating that CNT has lower selectivity to water than PDMS. Therefore, filling CNTs in the PDMS will improve the butanol/ethanol separation factor of the composite membrane, finally allowing the membrane to address the tradeoff between selectivity and permeability. Moreover, the computed adsorption energies of C4H9OH and C2H5OH onto the PDMS chain are higher than those in the external wall of CNT, indicating that these molecules are more favorable to adsorb on the PDMS chain, that is, in the CNT (closed-ended)/PDMS membrane, the molecules may be preferentially adsorbed and transported by the PDMS chain as they enter the composite membrane. In contrast, the limited free volume (several angstrom dimension) in the PDMS restricts the diffusion rate of penetrant molecules in the PDMS.13,14 Although the adsorption energy of external walls of CNTs is lower, the external walls of CNTs have more adsorption sites because of their higher specific surface area,50 which allows molecules to easily slide on the


11

Nano Letters 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

Page 12 of 24

smooth surfaces of the external walls, finally leading to a faster diffusion rate. Hence, filling the closed CNTs in the PDMS can increase the mass flux of the composite membrane. More strikingly, the adsorption energies of C4H9OH and C2H5OH onto the internal wall of the CNT are higher than those onto the external wall of the CNT and the PDMS chain. Accordingly, the molecules are more likely to be selectively adsorbed onto the smooth internal wall of the CNTs, thereby allowing rapid transport in the CNTs. The distribution of free volume in polymers can be used to interpret the diffusion coefficients of molecules in polymers. The diffusion coefficients of a given penetrant in the two structurally different polymers depend upon the fractional free volume.51 The channels inside the open-ended CNTs primarily act as a permeable inorganic phase to enlarge the free volume in the PDMS matrix; as a result, the diffusion coefficients of the aligned CNT (open-ended)/PDMS membrane are higher than those of the pure PDMS and the CNT (closed-ended)/PDMS membranes. As shown in Figure 4c, the open-ended CNTs play the most important role in facilitating the transportation of molecules in the membrane. In designing mixed-matrix membranes to overcome the permeability-selectivity tradeoff, the fillers characteristics and their homogeneous distribution in polymer are key factors. In principle, molecular sieving fillers (e.g., CMS and zeolites) often result in selectivity increase and permeability decrease.8,52,53 Fillers with interfacial voids, even when homogeneously distributed, often lead to decreased selectivity and increased permeability.8,54 Furthermore, molecular sieving fillers usually provide irregular diffusion paths when molecules pass through the inner channel (Figure S11). In addition, to understand the superiority of CNTs in the membrane, a similar “hamburger”-structured membrane with PDMS layers formed on top of porous PVDF substrate was fabricated for comparison, as shown in Figure S12. The CNT/PDMS membrane has higher mass flux and separation factors of alcohols than the PVDF/PDMS


12

Page 13 of 24 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

Nano Letters

membrane (Figure S13), which demonstrates that the orderly aligned CNTs are more beneficial than the irregular PVDF pores to provide preferable and selective pathways for molecules permeation. Therefore, when the aligned and open-ended CNT arrays are used as fillers, not only can the selectivity be increased via hydrophobic CNTs but the diffusion paths of molecules are also shortened via the orderly alignment of CNTs. Therefore, the aligned and open-ended CNT arrays can remarkably enhance both permeability and selectivity of the composite membrane. In conclusion, the aligned and opened CNT/PDMS membrane with a unique hamburger-like structure was successfully fabricated, and its pervaporation-ultrafiltration mechanism underlying penetrant molecules through the membrane was elucidated through membrane separation experiments and simulations. The annealed CNT membrane with high loading of 37 wt % CNTs showed a high total flux (1048.0 g m-2 h-1) and a butanol separation factor of 105.0 for n-butanol recovery in the VSVP process. The membrane does not suffer from the tradeoff between permeability and selectivity and allows both parameters to remarkably increase. The DFT computations indicate that the penetrate molecules diffusing through the CNT internal channels would be dominant in the membrane. In view of the versatile architectures of CNTs, this hamburger-like structured aligned CNT composite membrane exhibits excellent performance in biofuels separation and can be extended to applications in other fields, such as the chemical and environmental industries.

ASSOCIATED CONTENT Supporting Information


13

Nano Letters 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

Page 14 of 24

The Supporting Information is available free of charge on the ACS Publications website. The Supporting Information includes supporting figures and tables, details of material preparation, characterization, electrical and separation performance testing, a schematic illustration of the structure of the membrane, an equipment flowsheet, molecular diameters and the Hansen solubility parameters of solvents and PDMS, and representative membrane contrast (PDF). AUTHOR INFORMATION Corresponding Author * Email: [email protected] ORCID Chuang Xue: 0000-0002-3856-8457 Author Contributions ║

These authors contributed equally.

Notes The authors declare no competing financial interest. ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (NSFC) with Grant number (21576045 and 21878035), the Liaoning Innovative Talent Support Program (LR2017005), the Youth Science and Technology Star Project of Dalian (2017RQ003), the Talent Cultivation Plan of “Xinghai Scholar” from Dalian University of Technology, and the Fundamental Research Funds for the Central Universities (DUT16YQ103). We also thank Dr.


14

Page 15 of 24 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

Nano Letters

Gaohong He and Dr Xiaobin Jiang for their support on Young’s Modulus and gas permeation test. REFERENCES (1) Baeyens, J.; Kang, Q.; Appels, L.; Dewil, R.; Lv, Y.; Tan, T. Prog. Energ. Combust. 2015, 47, 60-88. (2) Cai, D.; Hu, S.; Chen, C.; Wang, Y.; Zhang, C.; Miao, Q.; Qin, P.; Tan, T. Bioresour. Technol. 2016, 220, 124-131. (3) Xue, C.; Liu, M.; Guo, X.; Hudson, P.; Chen, L.; Bai, F.; Liu, F.; Yang, S. Green. Chem. 2017, 19 (3), 660-669. (4) Khalife, E.; Tabatabaei, M.; Demirbas, A.; Aghbashlo, M. Prog. Energ. Combust. 2017, 59, 32-78. (5) Xue, C.; Liu, F.; Xu, M.; Tang, I.; Zhao, J.; Bai, F.; Yang, S. Bioresour. Technol. 2016, 219, 158-168. (6) Da Silva Trindade, W. R.; Dos Santos, R. G. Renew. Sust. Energ. Rev. 2017, 69, 642-651. (7) Xue, C.; Zhao, J.; Chen, L.; Yang, S.; Bai, F. Biotechnol. Adv. 2017, 35 (2), 310-322. (8) Park, H. B.; Kamcev, J.; Robeson, L. M.; Elimelech, M.; Freeman, B. D. Science 2017, 356 (6343), eaab0530. (9) Ong, Y. K.; Shi, G. M.; Le, N. L.; Tang, Y. P.; Zuo, J.; Nunes, S. P.; Chung, T. Prog. Polym. Sci. 2016, 57, 1-31. (10) Li, L.; Xiao, Z.; Tan, S.; Pu, L.; Zhang, Z. J. Membr. Sci. 2004, 243 (1-2), 177-187. (11) Wijmans, J. G.; Baker, R. W. J. Membr. Sci. 1995, 107 (1-2), 1-21. (12) Wijmans, J. G. J. Membr. Sci. 2004, 237 (1-2), 39-50. (13) Charati, S. G.; Stern, S. A. Macromolecules 1998, 31 (16), 5529-5535.


15

Nano Letters 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

Page 16 of 24

(14) Chang, K.; Chung, Y.; Yang, T.; Lue, S. J.; Tung, K.; Lin, Y. J. Membr. Sci. 2012, 417418, 119-130. (15) Okotrub, A. V.; Yudanov, N. F.; Aleksashin, V. M.; Bulusheva, L. G.; Komarova, O. A.; Kostas, U. O.; Gevko, P. N.; Antyufeeva, N. V.; Il'Chenko, S. I.; Gunyaev, G. M. Polym. Sci. Ser. A 2007, 49 (6), 702-707. (16) Okotrub, A. V.; Kubarev, V. V.; Kanygin, M. A.; Sedelnikova, O. V.; Bulusheva, L. G. Phys. Status. Solidi. B 2011, 248 (11), 2568-2571. (17) De Volder, M. F. L.; Tawfick, S. H.; Baughman, R. H.; Hart, A. J. Science 2013, 339 (6119), 535-539. (18) Spitalsky, Z.; Tasis, D.; Papagelis, K.; Galiotis, C. Prog. Polym. Sci. 2010, 35 (3), 357401. (19) Majumder, M.; Chopra, N.; Andrews, R.; Hinds, B. J. Nature 2005, 438 (7064), 44. (20) Coleman, J. N.; Khan, U.; Gun Ko, Y. K. Adv. Mater. 2006, 18 (6), 689-706. (21) Xue, C.; Du, G.; Chen, L.; Ren, J.; Sun, J.; Bai, F.; Yang, S. Sci. Rep-UK. 2015, 4 (1), 5925. (22) Zhang, Y.; Sheehan, C. J.; Zhai, J.; Zou, G.; Luo, H. Adv. Mater. 2010, 22 (28), 30273031. (23) Zhang, M. Science 2005, 309 (5738), 1215-1219. (24) Aliev, A. E.; J. Oh, M. E. K.; A. A. Kuznetsov, S. F.; A. F. Fonseca, R. O.; Lima, M. D.; Haque, M. H.; Gartstein, Y. N.; M. Zhang, A. A. Z. R. Science 2009, 323 (5921), 1575-1578. (25) Goh, P. S.; Ismail, A. F.; Ng, B. C. Composites, Part A 2014, 56, 103-126. (26) Hinds, B. J.; Chopra, N.; Rantell, T.; Andrews, R.; Gavalas, V. Science 2004, 303 (5654), 62-65.


16

Page 17 of 24 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

Nano Letters

(27) Holt, J. K.; Park, H. G.; Wang, Y.; Stadermann, M.; Artyukhin, A. B. Science 2006, 312 (5776), 1034-1037. (28) Baek, Y.; Kim, C.; Dong, K. S.; Kim, T.; Lee, J. S. J. Membr. Sci. 2014, 460, 171-177. (29) Lee, K.; Park, H. J. Membr. Sci. 2016, 501, 144-151. (30) Kim, S.; Fornasiero, F.; Park, H. G.; In, J. B.; Meshot, E.; Giraldo, G.; Stadermann, M.; Fireman, M.; Shan, J.; Grigoropoulos, C. P.; Bakajin, O. J. Membr. Sci. 2014, 460, 91-98. (31) Bui, N.; Meshot, E. R.; Kim, S.; Peña, J.; Gibson, P. W.; Wu, K. J.; Fornasiero, F. Adv. Mater. 2016, 28 (28), 5871-5877. (32) Hou, P.; Liu, C.; Cheng, H. Carbon 2008, 46 (15), 2003-2025. (33) Chi, Y.; Zhu, M.; Li, Y.; Yu, H.; Wang, H.; Peng, F. Catal. Sci. Technol. 2016, 6 (7), 23962402. (34) Hou, Z.; Cai, B.; Liu, H.; Xu, D. Carbon 2008, 46 (3), 405-413. (35) Ahn, K. S.; Kim, J. S.; Kim, C. O.; Hong, J. P. Carbon 2003, 41 (13), 2481-2485. (36) Cortese, B.; D, S.; Amone; Manca, M.; Viola, I.; Cingolani, R. Langmuir 2008, 24 (6), 2712-2718. (37) Wang, X.; Jiang, Q.; Xu, W.; Cai, W.; Inoue, Y.; Zhu, Y. Carbon 2013, 53, 145-152. (38) Georgakilas, V.; Bourlinos, A.; Gournis, D.; Tsoufis, T.; Trapalis, C. J. Am. Chem. Soc. 2008, 130 (27), 8733–8740. (39) Georgakilas, V.; Demeslis, A.; Ntararas, E.; Kouloumpis, A.; Dimos, K. Adv. Funct. Mater. 2015, 25 (10), 1481–1487. (40) Ma, P. C.; Kim, J.; Tang, B. Z. Carbon 2006, 44 (15), 3232-3238. (41) Wang, C.; Zhou, G.; Liu, H.; Wu, J.; Qiu, Y.; Gu, B.; Duan, W. J. Phys. Chem. B 2006, 110 (21), 10266-10271.


17

Nano Letters 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

Page 18 of 24

(42) Wang, R.; Shan, L.; Zhang, G.; Ji, S. J. Membr. Sci. 2013, 432, 33-41. (43) Jee, K. Y.; Lee, Y. T. J. Membr. Sci. 2014, 456, 1-10. (44) Li, S.; Srivastava, R.; Parnas, R. S. J. Membr. Sci. 2010, 363 (1-2), 287-294. (45) Zhu, C.; Chen, L.; Xue, C.; Bai, F. Biotechnol. Biofuels 2018, 11 (1), 128. (46) Bharadwaj, R. K.; Boyd, R. H. Polymer 1999, 40 (15), 4229-4236. (47) Majumder, M.; Chopra, N.; Hinds, B. J. ACS Nano 2011, 5 (5), 3867-3877. (48) Majumder, M.; Chopra, N.; Hinds, B. J. J. Am. Chem. Soc. 2005, 127 (25), 9062-9070. (49) Xue, C.; Liu, F.; Xu, M.; Zhao, J.; Chen, L.; Ren, J.; Bai, F.; Yang, S. Biotechnol. Bioeng. 2016, 113 (1), 120-129. (50) Song, J.; Gordin, M. L.; Xu, T.; Chen, S.; Yu, Z.; Sohn, H.; Lu, J.; Ren, Y.; Duan, Y.; Wang, D. Angew. Chem. Int. Ed. 2015, 54 (14), 4325-4329. (51) Jawalkar, S. S.; Aminabhavi, T. M. Polym. Int. 2007, 56 (7), 928-934. (52) Zimmerman, C. M.; Singh, A.; Koros, W. J. J. Membr. Sci. 1997, 137 (1-2), 145-154. (53) Xue, C.; Yang, D.; Du, G.; Chen, L.; Ren, J.; Bai, F. Biotechnol. Biofuels 2015, 8 (1), 105. (54) Bachman, J. E.; Smith, Z. P.; Li, T.; Xu, T.; Long, J. R. Nat. Mater. 2016, 15 (8), 845-849.


18

Page 19 of 24 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

Nano Letters

Figure legends Figure 1. Characterization of membrane fabrication. a) Fabrication process of the aligned and opened CNT/PDMS membrane; b) SEM image of CNT arrays; c) TEM image of CNT ends after Ar etching; d) TEM image of the cross-section of the membrane; e) SEM image of the external cross-section of the membrane edge after the Ar/CF4 plasma etching process; f) SEM image of the internal cross-section after adding a layer of PDMS (insets: the PDMS layer is indicated by a white arrow). Figure 2. CNT arrays and the membranes characteristics. a) Inner diameter distribution of the CNTs measured from a TEM image; b) comparison of the relation of tensile stress and strain with three types of membranes. Figure 3. The separation performance of four types of membranes for two typical bioalcohols separation. a) The total and constituent flux; b) the separation factor; c) the membrane thickness and performance for butanol recovery as reported previously and in our study. Figure 4. Schematic diagram of the molecule transport mechanism. a) The pure PDMS membrane; b) the aligned CNT (closed-ended)/PDMS membrane; c) the aligned CNT (openended)/PDMS membrane. Figure 5. a) Optimized structure of C4H9OH molecules onto the PDMS chain, the internal and external walls of the CNT; b) adsorption distance (d) and adsorption energy (Ea) of various molecules on different adsorption sites calculated from DFT simulations.


19

Nano Letters 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

Page 20 of 24

Figure 1.

Figure 2.


20

Page 21 of 24 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

Nano Letters

Figure 3.


21

Nano Letters 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

Page 22 of 24

Figure 4.


22

Page 23 of 24 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

Nano Letters

Figure 5.


23

Nano Letters 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

Page 24 of 24

TOC Graphic


24

Tuned Fabrication of the Aligned and Opened CNT ... - ACS Publications

Recommend Documents