Membrane Application on Porous Metallic Hollow ... - ACS Publications

May 30, 2017 - Positive thermal expansion coefficients (TECs) of 52 × 10–6 and 35 × 10–6 K–1 were experimentally calculated in the −116 to 2...
0 downloads 4 Views 2MB Size
Subscriber access provided by UNIV OF ARIZONA

Article

Structural Contraction of Zeolitic Imidazolate Frameworks: Membrane Application on Porous Metallic Hollow Fibers for Gas Separation Fernando Cacho-Bailo, Miren Etxeberria-Benavides, Oana David, Carlos Téllez, and Joaquin Coronas ACS Appl. Mater. Interfaces, Just Accepted Manuscript • Publication Date (Web): 30 May 2017 Downloaded from http://pubs.acs.org on May 31, 2017

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 free 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 accessible to all readers and 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.

ACS Applied Materials & Interfaces 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 30

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

Structural Imidazolate

Contraction Frameworks:

of

Zeolitic Membrane

Application on Porous Metallic Hollow Fibers for Gas Separation Fernando Cacho-Bailo,a Miren Etxeberría-Benavides,b Oana David,b Carlos Téllez,a Joaquín Coronasa,*

a

Chemical and Environmental Engineering Department and Instituto de Nanociencia de Aragón

(INA), Universidad de Zaragoza, 50018 Zaragoza, Spain b

Tecnalia Research and Innovation, Energy and Environmental Division, 20009 Donostia-San

Sebastián, Spain

* Corresponding author. E-mail: [email protected]¸ Phone: 0034 976 762471.

Abstract Positive thermal expansion coefficients (TEC) of 52·10-6 and 35·10-6 K-1 were experimentally calculated in the -116 – 250 ºC range for the III-phases of zeolitic imidazolate frameworks (ZIF) ZIF-9(Co) and ZIF-7(Zn), respectively, by means of the unit cell dimensions and volume of the materials in the monoclinic crystal system calculated from the XRD patterns. The unit cell dimensions and volume showed a significant expansion phenomenon as the temperature increased, by as much as 5.5 % for ZIF-9-III in the studied range. To exploit the advantages of

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

such thermal behavior, a new approach to the fabrication of ZIF-9-III membranes on thin, flexible and highly porous nickel hollow fiber (Ni HF) supports by a versatile and easy-controllable microfluidic setup is herein reported. These Ni HF supports result from the sintering of 25-µm Ni particles and display very positive mechanical properties and bending resistance. As compared to the traditional polymer-based HF membranes, the ZIF metal-supported membrane exhibited good durability and robustness throughout its operation in a wide temperature range and after heating and cooling cycles. These benefits derive from (1) the pore-plugging membrane configuration resulting from the high porosity of the support, and (2) the similarity between the TECs of the ZIF and the metallic support, both positive, which enhances their mutual compatibility. An increase in the H2/CO2 separation selectivity at low temperatures (as high as 22.2 at -10 ºC, along with 102 GPU permeance of H2) was achieved, in agreement with the structural variations observed in the ZIF material.

Keywords Metallic hollow fiber, MOF, membrane, gas separation, nickel

Supporting Information:

further fabrication parameters, XRD, TG, SEM and EDS

analyses on the green and sintered HF supports, ZIF powders and ZIF-9-III@Ni HF membrane. Lattice parameter data calculated under changes in temperature and the consequent TECs calculated for ZIF powders, and the equations used. Comparison of the H2/CO2 separation performance with other HF membranes reported in the literature.

1. Introduction Together with processes such as organic solvent nanofiltration,1 pervaporation2,3 and water purification,4 gas separation is one of the most important applications of membranes in chemical engineering today.5,6 A wide variety of oil-based polymers are available for membrane fabrication with reasonable raw material costs and ease of shaping, including polysulfones, polyamides and polyimides (such as fluorinated 6FDA-based or co-polyimide BTDA-TDI/MDI).2,7,8

2

ACS Paragon Plus Environment

Page 2 of 30

Page 3 of 30

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

Nevertheless, polymeric materials possess an intrinsic permeation/selectivity upper bound that limits their further development. MOFs (metal organic frameworks), porous hybrid materials formed by metal clusters joined by organic ligands, are promising materials for complementing membranes.9-11 Their potential in gas separation is determined by their narrow porosities, chiefly those of the so-called ZIFs (zeolitic imidazolate frameworks) with pore apertures below 5 Å.12,13 There are numerous reports concerning the addition of ZIFs, and MOFs in general, to polymers as fillers and as supported layers applied to the separation of gas mixtures with positive results.1417

In terms of membrane configuration, hollow fibers (HF) are extensively used in gas separation processes due to their high degree of efficiency and compactness given by their high area to volume ratios (in the range of 103-104 m2 per m3), their scalability and their enhanced mass transport resulting from their good contact with the feed stream.18,19 Recently, some procedures have been reported for MOF selective layer deposition onto porous polymeric hollow fiber supports.20,21 Although a high degree of compatibility is observed between MOF and polymeric membrane supports because of their common organic character,22-24 differences in their thermal and operational (e.g. annealing and plasticizing) stability may prohibit their use in a wide range of conditions. In addition, the presence of metal ions may cause a catalytic effect in the polymer degradation.24 The solution might then be the use of HF-sized porous inorganic supports with suitable flexibility, porosity and robustness. Related to this, some studies of HF-shaped porous metallic supports made of titanium,25 nickel26 and stainless steel27 have been reported. Furthermore, the individual thermal expansion coefficients (TEC) of the components forming the membrane play a key role in the thermal performance of the resulting composite membrane. Similar TECs would produce good compatibility between the layered MOF and the support and consequently enhanced membrane durability under temperature fluctuations. From the theoretical prediction made by Bouëssel du Bourg et al.,28 as shown in Figure 1 and Table S1, both sod-ZIFs and metals display positive TECs. In contrast, zeolites and carboxylate-based MOFs may have negative TECs.29-35 In fact, zeolite crystals show an irregular thermal expansion and contraction

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

behavior depending on the water content and the temperature range tested, as reported for LTA-, FAU-, MOR- and MFI-type zeolites commonly used to prepare zeolite membranes.36 This is the reason for the loss of performance exhibited by ceramic-supported zeolite membranes under temperature variations, affecting the zeolite crystals chemically bonded to the support, unless a suitable calcination temperature program is applied.37 Some other metal-organic frameworks with large positive and negative TECs,38,39 including Ag-methylimidazolate as studied by Ogborn et al.,40 are included for comparison. The averaged TECs in Figure 1 do not show anisotropic disparities between the directions or variations in narrow temperature ranges.

Figure 1. Linear thermal expansion coefficients (TEC) of materials suitable for use in the fabrication of membranes, see Table S1.28-30,32-34,38-40 Zeolites and carboxylate-based MOFs generally display negative TECs. Simulation of imidazole-based sod-ZIFs predicted positive TECs similar to those of metals such as stainless steel (SS) or Ni and lower than those of polymers.

The separation of H2 from CO2 is of great importance for the development of technologies for CO2 capture and storage (CCS) and hydrogen purification.41-44 However, this separation is challenging because of its high operational temperature and pressure. ZIF-7 (Zn-bIm, zincbenzimidazolate, with a sod topology) has previously been proven to act as a molecular sieve for CO2 molecules. Li et al. achieved a H2/CO2 selectivity of 13.6 with a ZIF-7 flat-supported membrane operated up to 200 ºC.45 Also, ZIF-9 (isostructural cobalt-bIm) was combined with ZIF-8 and ZIF-67 in double-layered membranes that improved their performance in H2/CO2 separation.46,47 The phases I and II (also known as large- and narrow-pore, respectively) in

4

ACS Paragon Plus Environment

Page 4 of 30

Page 5 of 30

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

benzimidazolate ZIFs display a high CO2 adsorption capacity.46 However, the dense lamellar phases (also known as phase III) of these materials, driven by the atom rearrangement in water media,48,49 exhibit a low CO2 adsorption capacity50 and more constricted pores, thus anticipating an enhancement in H2/CO2 separation. Taking advantage of this, Peng et al. achieved a H2/CO2 separation selectivity of 291 by means of an oriented deposition of exfoliated ZIF-7-III sheets on a flat support.51 Herein, thermal expansion coefficients were experimentally calculated for ZIF-7-III and ZIF-9III. To the best of our knowledge, these are the first experimental TEC values for a ZIF material and they set the stage for the study of the permselective separation properties of a ZIF-9-III-based membrane in a wide temperature range. A flexible and highly porous metallic HF support with an outer diameter of 610-µm was developed from 25-µm Ni particles. It was then used as a support for bIm-ZIF (ZIF-9-III) growth with a high potential in H2/CO2 separation. The good mechanical properties of the raw support together with the similarity of the expansion phenomena of ZIFs and Ni with temperature (Figure 1 and Table S1) suggest suitable stability and durability. The synthesis of a ZIF on a support with a higher thermal stability (as compared to common polymeric hollow fibers) than the ZIF itself would allow extracting the maximum benefit from the exceptional performance of such MOFs at relatively extreme temperatures.

2. Experimental Section 2.1 Metallic support fabrication Ni powder with an average particle size of 25 µm was purchased from TLS Technik GmbH & Co. (Germany). Polyethersulfone (PES, Ultrason E 6020, BASF) was used as polymer and Nmethylpyrrolidone (NMP, 99 wt%, Sigma Aldrich) as solvent. Prior to use, the PES was dried in a vacuum oven at 100 °C for 12 h. All the other chemicals were used without further treatment. The fabrication of Ni precursor tubes was carried out using a dry wet spinning process.25,27 The dope mixture contained ~80 wt% metal powder, polymer, NMP and other additives and was

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

prepared as follows: (1) the PES was dissolved in NMP in a glass vessel for 24 h on a roller tube mixer; (2) metal powder was added to the vessel and stirred with a mechanical mixer for 60 min at 350 rpm. The vessel remained sealed on the roller overnight and was heated at 30 – 40 ºC by an IR lamp. The dope was loaded into a syringe pump and degassed for 24 h at room temperature before spinning. Ni green compact hollow fibers were formed when the bore and dope fluids were co-extruded through a spinneret at a controlled flow rate and passed through a coagulation water bath. The main spinning parameters are listed in Table S2. The thermal treatment was carried out by means of a tubular furnace (Termolab T.H.) at a controlled heating rate of 5 ºC·min-1. The removal of the polymer was performed at 600 °C for 60 min in air and Ni particles were sintered at 800 ºC for 120 min in inert atmosphere.

2.2 ZIF-9 @ Ni HF membrane fabrication A 13 cm long Ni HF support was first sealed to the 1/8 inch stainless-steel (SS) module by means of an epoxy resin, in the same gas permeation module where the membrane is subsequently tested for mixture separation. For membrane fabrication, the module was connected to the microfluidic setup 1/16 inch tubing, whose description can be found elsewhere,21,52 by means of a 1/16-1/8 inch SS-reducer. Because of its enormous initial porosity, the Ni support was immersed in pure PEG400 in order to confine the MOF growing in the HF lumen. About 5 mL PEG400 were dropped externally surrounding the Ni HF support. Metal and ligand containing solutions, contained separately in their corresponding syringes, were alternately injected inside the HF lumen using microfluidic pumps (NE-300, New Era Pump Systems). This was done in a sequence of 6 steps of 2 mL each (6 mL each of the metal and ligand solutions, totaling 12 mL) at a 40 µL·min-1 flow rate. Absolute ethanol was injected for soaking (2 mL) and washing (5 mL), and also between each metal-ligand solution change (0.2 mL each). For ZIF-9, Co(bIm)2, the metal solution contained 0.50 g cobalt acetate tetrahydrate (Co(CH3COO)2·4H2O, reagent grade, Sigma-Aldrich) and 0.27 g ammonium hydroxide solution (NH4OH, 28–30% NH3 basis, Sigma-Aldrich) in 10 mL absolute ethanol,

6

ACS Paragon Plus Environment

Page 6 of 30

Page 7 of 30

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

whereas the ligand solution contained 0.24 g benzimidazole (bIm, 98%, Sigma-Aldrich) in 10 mL absolute ethanol. Subsequently, a step involving the ZIF-9 conversion into its dense form (ZIF-9-III) was performed. 10 mL of deionized water was injected at 40 µL·min-1 into the HF membrane lumen. The membrane outer surface was abundantly washed with deionized water to remove the remaining PEG400, and the whole assembly was left to dry at room temperature overnight. Outer and inner PDMS (polydimethylsiloxane) coatings were applied for membrane healing. A 3 wt% Sylgard-184 solution (a polymer and curing agent mixed in a 10:1 proportion, Sigma-Aldrich) in hexane was poured in contact with the outer membrane surface for 5 and 15 min, while injecting air inside the inner lumen of the membrane using the microfluidic pumps. The PDMS solution was then removed, while the membrane was dried under ambient conditions for at least 1 h. An inner-surface PDMS coating was also applied. 0.5 mL of PDMS solution was injected inside the membrane lumen at a 40 µL·min-1 flow rate and removed by air-injection immediately afterwards. The PDMS-double-coated membrane was then cured at 100 ºC overnight. The microfluidic flow conditions during the membrane fabrication (40 µL·min-1) provided a strictly laminar flow regime with no turbulence inside the HF lumen, where the ZIF crystallization took place. A Reynolds number of 1.5 was calculated. Microfluidics therefore turned out to be a versatile and easily-controllable technique for MOF-supported membrane fabrication, chiefly when applied on HF shaped supports.21

2.3 Characterization The green compact and sintered Ni HF supports were characterized by scanning electron microscopy (SEM, FEI Inspect F50). SEM images were taken at a voltage of 5-15 kV with specimens coated with platinum under vacuum conditions. Composition maps of the cross sections were analyzed by energy-dispersive X-ray spectroscopy (EDS) using a Quanta FEG-250 SEM microscope. Pore size and distribution, surface area, density and porosity were measured using Hg porosimetry (Micromeritics AutoPore IV 9500 V1.09). N2 permeation was measured in

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

a pressure controlled dead-end permeation setup at 2 bar and room temperature. The permeation module contained one fiber with a 5-10 cm active length. N2 permeance was calculated as the ratio between the gas flux per unit area and the trans-membrane pressure difference. The ZIF-9 powder was separated by centrifugation at 10,000 rpm from the metal and ligand containing solutions collected during the membrane fabrication. It was washed with absolute ethanol and then dispersed in deionized water for 4 h, in an identical manner as with the MOF in the membrane. The dispersion was centrifuged again and dried at 100 ºC. For comparison, ZIF-7 powder was synthesized in bulk in a similar manner to that used for ZIF-9, replacing the cobalt source with 0.44 g zinc acetate dihydrate (Zn(CH3COO)2·2H2O, reagent grade, Sigma-Aldrich) in the metal solution. Powder XRD (X-ray diffraction) patterns were obtained using a Rigaku D-Max diffractometer with a rotating Cu Kα anode and a graphite monochromator (λ=1.542 Å) operating at 40 kV and 80 mA, from 2·theta 4 to 40 º with a 0.03 º step·s-1. Diffractograms at temperatures from -116 to 250 ºC were obtained in a vacuum atmosphere making use of an attachment for medium and low temperature. After cooling down, an extra XRD analysis was done at room temperature to check the powder crystallinity and to discard any possible contribution from molecule loading effects. Unit cell dimensions (a, b and c) and volume (V) were numerically inferred at each temperature from the d-spacings of the most intense diffraction planes in the experimental XRD data, namely, (0 0 2), (2 2 1), (-2 -2 2), (2 2 3) and (-2 2 4) in the reported crystallographic structures,48,49 by minimizing the least-squared residue in Equations S1 and S2 (included in the Supporting Information), while the β angle and the structural parameters (crystal system and space group) were maintained fixed. This procedure allowed the estimation of an error of around 0.6 % in the calculation of the cell dimensions (when compared with the existing data obtained at room temperature), much lower than the variations observed in the unit cell lengths as a function of temperature. These equations are valid for a monoclinic crystalline system, such as those studied for ZIF-9-III and ZIF-7-III, relating the d-spacing (d) in the XRD patterns with the (h k l) Miller indices of every plane. Linear (α) and volumetric (αv) TECs were calculated for several

8

ACS Paragon Plus Environment

Page 8 of 30

Page 9 of 30

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

temperature ranges along each axis (see Equations S3 to S4), assuming orthogonal coordinates despite the monoclinic system (Equation S5).34,35 SEM images were obtained using a FEI INSPEC-F. Elemental analyses by EDS were carried out with an INCA PentaFET x3 (Oxford Instruments). TG (thermogravimetric) analyses were obtained with a METTLER Toledo TGA/DSC SF/755 from 25 to 750 ºC at a rate of 10 ºC·min-1. Images showing the ZIF structures were rendered using Diamond Crystal Impact software.53 The ZIF-9@Ni HF membrane was tested for gas mixture separation using the same module in which it was sealed for the MOF synthesis. Separation tests were carried out using the WickeKallenbach technique described elsewhere.21,52 Equimolar gas mixtures were fed through the HF membrane lumen while the permeate stream was swept with Ar and analyzed in an Agilent 3000A gas micro-chromatograph. Fluxes in the 10-25 cm3(STP)·min-1 range were fed. A total pressure of 1.3 bar was applied on both membrane sides, the total pressure drop through the membrane being null. An effective permeable area of 1.39 cm2 was considered. Permeance values (mol·m2

·s-1·Pa-1) were calculated using the log-mean partial pressure difference along the HF length,

whereas the mixture separation selectivity corresponded to the ratio between permeances.21 H2/CO2 and He/CH4 mixtures were analyzed at temperatures from -10 to 250 ºC. The membrane module was heated in an oven and cooled in ice/water or ice/acetone baths.

3. Results & Discussion 3.1 Benzimidazolate-ZIF experimental TEC calculation Both as-made ZIF-7 and ZIF-9 are also named I-phases, while II- and dense, layered III-phases can be obtained by means of thermal and water treatments, respectively.48,50 The powdered materials showed XRD patterns corresponding to the III-phases of ZIF-9(Co) and ZIF-7(Zn), respectively (Figure S1).49,50 The transitions from the as-synthesized I-phases to these III-phases were obtained during the densification step with deionized water sequentially added in the synthesis procedures. CCDC numbers/codes 675375/KOLYAM49 and 988184/SOTYIL48 of the

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

structures of ZIF-9-III (Co-bIm) and ZIF-7-III (Zn-bIm), respectively, led to the prediction of promising H2/CO2 separation by molecular sieving through the 0.21 nm-diameter pore along the “c” direction.51 In order to reveal the ZIF behavior as a function of temperature, the evolution of the XRD patterns of ZIF-9(Co) and ZIF-7(Zn) powders was followed in the -116 – 250 ºC range (Figures 2a, S2 and S3). Unit cell dimensions and volume shifts with respect to the temperature were calculated by minimizing the least-squared distance of the d-spacing data in the experimental XRD patterns corresponding to the monoclinic crystal system of ZIF-7-III and ZIF-9-III (Equations S1 and S2) at each temperature.48,49 Values of the d-spacing of the five most intense peaks (the other nonconsidered reflections also corresponding to the sod structure) in the diffractograms were taken for calculation. Figure 2b and Tables S3 and S4 show the cell parameters obtained at each temperature for ZIF-9-III and ZIF-7-III, respectively, and the percentage variation with respect to the dimensions of the unit cell at ambient temperature. Averaged experimental linear TECs along each direction and in the high and low temperature range were calculated with Equations S3 to S5,34,35,54 and are shown in Tables 1 and S5.

10

ACS Paragon Plus Environment

Page 10 of 30

Page 11 of 30

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 2. Normalized XRD patterns of ZIF-9-III as a function of temperature in the -116 – 250 ºC range (a). No phase transition but a clear shift in the diffraction peaks was observed, related with a structure expansion/contraction phenomenon. ZIF-7-III patterns are shown in the Supporting Information (Figure S3) for similar behavior. The Miller indices of the peaks used in the equations for unit cell calculation are indicated. ZIF-9-III and ZIF-7-III unit cell dimensions (in the a, b and c directions) and volume dependence with temperature were calculated from the experimental XRD patterns considering a monoclinic crystalline system (b). See Supporting Information and Experimental for further details.

An overall positive linear TEC (α=52.2·10-6 K-1, calculated from the volumetric TEC) was found for ZIF-9-III. However, a significant anisotropy in the thermal expansion was observed,55 as well as a different behavior in the studied temperature ranges. The variations observed in the a- and caxes led to positive TECs (Table 1), in contrast with the calculated negative TEC αb=-143.0·10-6 K-1, related to a contraction of the material along the b-axis with increasing temperature. Regarding the TEC trend with temperature, about 3-fold larger variations were found in the ZIF9-III structure below room temperature (92.6·10-6 K-1 TEC in the -116-20 ºC range) compared to

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

Page 12 of 30

the 20-250 ºC range (where a TEC of 27.4·10-6 K-1 was calculated). The experimental averaged TEC for the ZIF-9-III fell above that theoretically predicted for a sod-Zn(Im)2 (8.3·10-6 K-1)28 and was also larger than that of the metal nickel (13.3·10-6 K-1), a priori chosen as an appropriate membrane support in this work (see Figure 1). In any event, both the ZIF-9-III and the nickel support display positive TECs, in contrast with the general behavior of carboxylate-based MOFs and zeolites. Similar conclusions were found in the study carried out on the ZIF-7-III powdered material, for which a TEC of α=34.5·10-6 K-1 was calculated (see ESI, Table S5). The same trends in the unit cell lengths and volume were observed as those for ZIF-9-III, indicating a high anisotropy in the b-axis compared to the a- and c-axes. Also, a larger expansion response of the MOF structure to temperature variations was observed at temperatures below 20 ºC (Figure 2b and Table S4). Table 1.Linear thermal expansion coefficients (TEC) obtained for ZIF-9-III.

-116 ºC – Troom (20 ºC) Troom (20 ºC) – 250 ºC -116 ºC – 250 ºC

αa ·10-6 a [K-1] 255.0 127.7 177.8

αb ·10-6 a [K-1] -135.1 -150.9 -143.3

αc ·10-6 a [K-1] 160.0 110.3 130.3

a

Linear TEC in the corresponding direction, see Equations S3 to S5.

b

Averaged TEC; α = αv/3 (αv = volumetric TEC).

α ·10-6 b [K-1] 92.6 27.4 52.2

Figure 3 shows a rendering of the ZIF-9-III structure from two different viewing directions. The dense phase under study showed a laminar conformation with the material sheets placed along the c-axis direction bonded by weak van der Waals forces.51 From the trends observed in the unit cell lengths and volume with temperature (Figure 2b), an increase in the inter-laminar spacing (cdirection) and overall cell volume within the entire temperature range was predicted, as much as 5.5 % for ZIF-9-III. The β-angle was considered constant for the calculation process in the studied temperature range. In addition, an inverse expansion-contraction behavior along the perpendicular a- and b-dimensions, respectively, can be deduced (the b-axis length contracted 3.5 % while the a-axis expanded 2.9 % at 250 ºC compared to 20 ºC, see Table S3). As stated before,40 a strong anisotropy with coupled positive and negative TECs would be typical of materials with hinge-

12

ACS Paragon Plus Environment

Page 13 of 30

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

connected building units with very soft additional interactions, as appears to occur in the ZIF-7III and ZIF-9-III structures.

Figure 3. Rendering of the ZIF-9-III structure from two different viewing directions. Hydrogen atoms are hidden for clarity. The material layers along the c-direction can be clearly distinguished, as well as the 4-member ring pores in each layer (a-b plane). Figure S4 in the Supporting Information shows the rendering of the ZIF-7III structure.

This behavior of the structure suggested an enhanced performance in molecular-sieved gas separation at low temperatures. To gain insight into a real application, a ZIF-9-III membrane was fabricated, characterized and tested for H2/CO2 separation over a wide range of temperatures. Compared with ZIF-7-III, ZIF-9-III exhibited a larger TEC and more constricted pores due to the shorter Co-N than Zn-N distances. The porous Ni HF used as support provided the appropriate mechanical and thermal stability and durability for the membrane during the heating and cooling cycles.

3.2 ZIF-9-III membrane on Ni HF flexible support Figures 4 and S5 show SEM cross-sectional images of the green and sintered Ni hollow fibers. They showed good circularity and concentricity with a macroporous spongy-like structure for the green support (Figure S5). The Ni particles were distributed uniformly within the polymer matrix with a remarkably high loading. Different magnifications of the cross sections of the sintered fibers and the outer fiber surface are shown in Figures 4a and 4b, respectively. The outer diameter (610 µm) and wall thickness (130 µm) of the sintered fibers was slightly lower than that of the precursor green fiber. This low cross sectional shrinkage was related to a high particle loading in

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

the precursor fiber. It can be observed that the sintering process gave rise to a structure composed of well interconnected metal particles. Figure 4b shows that the fiber breakage, when obtaining the specimen for characterization, occurred at the bridges formed between metal particles during sintering. Table 2 and Figure S6 show some structural and textural properties of the sintered Ni HF supports and the pore size distribution obtained by Hg intrusion porosimetry. Table 2. Physical, textural and permeation properties of the sintered Ni HF.

N2 permeance Total pore area Modal pore diameter Open Porosity Bulk Density Apparent density

1.8·10-6 0.9 5.2 23.4 4.7 6.1

mol·m-2·s-1·Pa-1 m2·g-1 µm % g·mL-1 g·mL-1

Micrometer-sized cubic crystals with rounded edges and showing the exfoliation planes were observed in the SEM images of the ZIF-9-III powder (Figure S7). Consistent with the XRD pattern of the powdered material, neither a significant exfoliation nor a preferential growth could be inferred in the ZIF-9 material in the fabricated membrane.51 Micrometer-sized ZIF-9 crystals in the 1-5 µm range were observed, suitable for plugging the voids left by the Ni particles in the sintered HF support (5.2 µm, see Table 2 and Figure S6). However, high particle size polydispersity in the powdered ZIF-9-III sample was expected, because the sample was collected throughout the duration of the membrane fabrication. Besides, ZIF-9-III degraded at 365 ºC (obtained as the maximum in the derivative of the weight loss curve, Figure S8) as indicated by the TG analyses in air. The ZIF-7-III TG curve is also shown for comparison, decomposing at a higher temperature (585 ºC). Therefore, ZIF-9-III(Co) showed a 220 ºC-reduced thermal stability in comparison with ZIF-7-III(Zn) because of the greater catalytic effect of cobalt in the MOF material oxidation reaction as compared with that of zinc. However, ZIF-9-III is stable at temperatures up to 600 ºC in a non-oxidative atmosphere such as that of the gas separation tests.56 The TG curve of the entire membrane in an air atmosphere is shown in Figure S8, from which a 4.1 wt% of ZIF-9-III content in the membrane can be calculated. Compared with previously reported MOF@polymeric HF (with MOF contents in the 15-20 wt% range),17,21 this Ni HF

14

ACS Paragon Plus Environment

Page 14 of 30

Page 15 of 30

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

membrane contained a lower MOF loading due to the high density of the support. Supported-ZIF9-III now degraded at a higher temperature (404 ºC) than the powder because of its shielded integration within the nickel particles.

Figure 4. SEM images of the sintered porous Ni HF (a,b) and the fabricated ZIF-9@NiHF membrane (c-h). Cross section and wall detail (a) and outer surface (b) of the sintered Ni HF at different magnifications. ZIF9@NiHF membrane cross section images at two different magnifications (c,d). Yellow shows the metal Ni presence from the support (e), whereas red shows Co (f,g) and green shows Si (h), coming from the PDMS healing coating the inner surface and some remaining voids in the membrane. A ZIF-9(Co) MOF growth through the entire thickness was observed because of the high porosity of the Ni HF support, resulting in a poreplugging membrane.

The microscopic observation of the ZIF-9@NiHF membrane showed MOF particles filling the inter-particle gaps of the support through the entire thickness, thus filling the support pores (Figure 4c-d), in a similar manner to the pore-plugging MFI type zeolite membranes on alumina described by Miachon et al.57 EDS mapping showed the highly homogeneous presence of the Co from ZIF-9-III in every spot in the membrane, being completely mixed with Ni from the support (see Figure 4e-h). An occupation of 90 % of the initial porosity of the support (21.0 % of the entire membrane volume, see Table 2) was estimated for the ZIF-9-III considering its structure

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

Page 16 of 30

density (1.39 g·mL-1)48 and an 8.9 g·mL-1 density for Ni. The remaining volume fraction continued to be void or occupied by the PDMS coating. In this respect, the Si-EDS mapping showed how the PDMS not only continuously coated the membrane inner surface but also filled some gaps in the membrane thickness. The ZIF-9@Ni HF membrane was then tested in the separation of the gas mixtures of concern,5 namely, H2/CO2 and He/CH4. The membrane showed separation selectivities above the corresponding Knudsen values for these two mixtures (Table 3 shows the permeation results at 35 ºC). A ZIF-9-III micropore sieving effect can thus be deduced, the larger molecules CO2 (3.3 Å) and CH4 (3.8 Å) being preferentially excluded by the ZIF limiting pore diameter.51 As H2 (2.9 Å) and He (2.6 Å) have larger kinetic diameters than 0.21 nm, some flexibility must be assumed in the bonds of the ZIF structures.58,59 Also, the gas permeances were influenced by a molecular weight-dependent Knudsen diffusion so that CO2 permeance was ca. 0.6 times that of CH4 (Figure S9). He and H2 permeance values followed the same trend, the lighter H2 permeating faster than the smaller He. The H2/He separation then exhibited a promising selectivity value (1.3, 1.4 at -10 ºC, also consistent with the Knudsen selectivity of the mixture). This behavior has been previously reported for other MOF supported membranes, i.e. those of MOF-5 and HKUST-1,15,60 and is in contrast with the permeance values obtained in earlier reported ZIF membranes with non-laminar structures,17,45 where the gas permeances were related to the molecular kinetic diameters. PDMS in the membrane was assumed to provide a high permeance with respect to the ZIF with a nonsignificant (below Knudsen) selectivity in the H2/CO2 mixture separation.61 Table 3. Mixture permeation data obtained with the as-synthesized ZIF-9@Ni HF membrane at 35 ºC.

Gas

a

He H2 CO2 CH4

Kinetic diameter [Å] 2.6 2.9 3.3 3.8

Permeance [mol·m-2·s-1·Pa-1] [GPU] -8 1.6·10 48 2.3·10-8 64 1.5·10-9 4.5 2.1·10-9 6.2

/He 1.3a

Ideal selectivities calculated from the gas permeances of the studied mixtures.

16

ACS Paragon Plus Environment

Selectivity /CO2 /CH4 a 10.6 7.8 14.1 10.4a 0.7a

Page 17 of 30

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, an increase in the gas permeances was further observed during the membrane activation process carried out in situ by means of several heating-plateau-cooling cycles up to 250 ºC while separating the H2/CO2 mixture using Ar as a sweep gas (see Figure S10). The H2 permeance increased from 2.2·10-8 to 4.8·10-8 mol·m-2·s-1·Pa-1 at 35 ºC in the resulting stationary state, while the selectivity remained almost unchanged (13.2). This stabilization in the gas permeances can then be related to an increase in the support porosity and not with changes or activation of the supported MOF. Two possible reasons were hypothesized: (1) the evaporation of residual PEG400 (PEG400 starts evaporating at 200 ºC, see Figure S11), water or solvent remaining after the MOF synthesis, and (2) the partial reduction of NiO, found in low amounts in the initial support, under a reductive atmosphere (H2/CO2/Ar) during the membrane testing. NiO reduction into Ni might increase the free volume and therefore lead to a higher gas flow through the membrane. This was first checked via TG analysis (Figure S12). The membrane testing conditions were replicated, a sample of the as-made Ni HF support being maintained at 250 ºC for 24 h under 50/50 v/v H2/CO2. No mass loss was observed, even though the superficial EDS analyses positively showed a decrease in the oxygen content from 4.5 wt% to 1.3 wt% in the final sample with respect to the initial Ni support (see Table S6). This oxide content observable on the surface (but not in the bulk, see Figure S13) may have caused the reduction in the measured apparent density of the HF support (6.1 g·mL-1, Table 2) with respect to that of pure Ni (8.9 g·mL-1). On the other hand, PDMS coating is stable up to 300 ºC and is not considered to have changed during the membrane operation. ZIF structural contraction in gas separation performance The H2/CO2 separation was then studied as a function of temperature in the -10 – 250 ºC range (Figure 5 and Table S7). A clear decrease in selectivity with temperature was observed: a plateau at around 6.9 was reached for temperatures above 200 ºC whereas the highest selectivity (22.2) was achieved at -10 ºC with an ice/acetone cooling bath. The apparent activation energy calculated for CO2 (6.2 kJ·mol-1) was three times higher than that calculated for H2 (2.1 kJ·mol-

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

1

), related with the observed decrease in separation selectivity with temperature (see the

corresponding Arrhenius plots in Figure S14).

Figure 5. Permeation data dependence on temperature of the activated ZIF-9-III@Ni HF in the H2/CO2 mixture separation. A decreasing trend in the H2/CO2 separation selectivity (dotted lines shown as a visual guide) with temperature was observed.

These results are in contrast with the behavior of traditional ZIF and zeolite membranes, where the CO2 adsorption and condensation in the micropores is enhanced at low temperatures, giving rise to low or even reversed (below 1) H2/CO2 separation factors.62 In this case, a very low CO2 adsorption in the ZIF-9-III material, as reported by He et al. for ZIF-7-III,50 contrary to the high CO2 adsorption capacities of I/II-phase bIm-ZIF materials,46,63 would cause this increase in H2/CO2 selectivity at low temperatures. As a probe, a supported SIM-1 membrane on P84® HF was tested for H2/CO2 in the -10 to 35 ºC range, displaying a maximum in the CO2 permeance at temperatures around 0 ºC and a CO2-favorable 0.25 H2/CO2 separation selectivity (see Figure 6) derived from the above-mentioned CO2 adsorption. The H2 permeance, hardly adsorbed, linearly increased with temperature. A relatively high CO2 adsorption capacity has been previously reported for SIM-1 (Zn(4-methyl-5-imidazolcarboxaldehyde)2, also known as ZIF-94), a ZIF material containing an aldehyde group.64 This observed dependence of the gas separation performance on temperature (displaying an enhanced separation selectivity at low temperatures) could be directly related to the positive TEC

18

ACS Paragon Plus Environment

Page 18 of 30

Page 19 of 30

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

previously calculated for ZIF-9-III. The reduction in the unit cell volume and pore constriction would increase the CO2 molecular sieving from H2 at low temperatures (Figure 5). It would also correlate with the 3-fold higher apparent activation energy in CO2 permeance, once the CO2 adsorption in ZIF-9-III can be considered negligible. This behavior might also cause the trend observed by Wang et al. when testing a ZIF-100 membrane on an α-Al2O3 disk, this ZIF being based on the 5-chlorobenzimidazole ligand:65 the H2/CO2 separation selectivity decreased from 72 at 25 ºC to about 20 at 150 ºC. This material displays a typical XRD pattern of a laminar material, similar to that of ZIF-9-III, and also a low CO2 adsorption capacity (0.9 mmol·g-1).

Figure 6. Performance of a SIM-1@P84® membrane in the H2/CO2 mixture separation at low temperatures. A maximum in the CO2 permeance at 0 ºC and selectivities below 1 can be observed as a result of the CO2adsorption in the aldehyde-containing SIM-1 material. In contrast, the ZIF-9-III displayed negligible CO2 adsorption (see performance data of the ZIF-9@Ni HF membrane in H2/CO2 separation in Figure 5 for comparison).

To the best of our knowledge, the H2/CO2 separation selectivity of 22.2 (at the 102 GPU permeance of H2 at -10 ºC) is the highest ever obtained with a MOF membrane supported on an inorganic hollow fiber (see Figure S15 and Table S8 for bibliographic comparison). To date, almost every attempt using MOF-supported tubular or HF membranes has been made with ZIF8. A higher H2/CO2 selectivity (32.4) has been reported only in the study by Li et al. describing a NH2-MIL-53 membrane supported on a PVDF polymeric HF.15 Hollow fibers are preferred over flat membranes due to their industrial interest resulting from their feasibility for up-scaling. Also,

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

their inorganic (metal) composition would allow maximum advantage to be made of the high thermal stability of ZIFs. Figure S15 shows a bibliographic comparison regarding HF membranes in H2/CO2 separation, the permeance values being normalized by the support outer diameter (OD) as a measure of their efficiency. Smaller membrane diameters (HF OD generally considered below 1 mm) provide high area-to-volume ratios and therefore a positive degree of intensification in the separation process. Our approach is an answer to the challenge of conforming thin inorganic HF supports that have similar performance to polymeric HFs in terms of separation efficiency, operation stability and mechanical features. The Ni HF supports developed in this work, with an OD of 610 µm, displayed more than 6500 m2·m-3. Nevertheless, this H2/CO2 separation selectivity of 22.2 falls below those achieved with delaminated sheets of ZIF-7-III material (291) conforming a flat membrane.51 In that work, an inhibition of the re-stacking of the material plates was revealed as determinant for avoiding interlaminar gas diffusion, leading to a significant improvement in the separation performance. In the absence of inter-laminar pores, gas molecules were forced to pass only through the highly restrictive 4-membered-ring pores, as shown in Figure 3. In the in situ fabricated ZIF-9@NiHF membrane developed here, neither a non-preferential growth nor a delamination of the cobaltbased ZIF-9-III crystals can be expected. Therefore, there was a likely presence of gas flux through the interlaminar space (along directions “a” and “b” in Figure 3), where a lower molecular sieving effect can be expected. However, an interesting property was revealed: the contraction of the ZIF-9 structure which, together with an absence of CO2 adsorption in the material, led to an improvement in the H2/CO2 separation performance at low temperatures. Finally, this high separation factor, together with the SEM and EDS analyses and the ZIF powder collected during the membrane synthesis, indirectly evidenced the MOF presence in the membrane. Furthermore, a wide range of operational temperatures was available for testing, from -10 ºC to 250 ºC, measured through a significant number of heating and cooling cycles (see Figure S10). This reliable membrane performance can be related to the compatibility of the materials in terms of thermal properties. It is also worth mentioning that the membrane was tested for about one

20

ACS Paragon Plus Environment

Page 20 of 30

Page 21 of 30

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

month of accumulated operation, including tests with several gas mixtures and even at temperatures below 0 ºC, while the Ni support retained its quality and the adhesion with the MOF remained. The ZIF@Ni HF membrane withstood the volume variations in the MOF structure (as high as 5.5 % in the entire temperature range, see Table S3) and displayed a good robustness in testing, maybe because of its pore-plugging conformation. This type of membrane architecture was thoroughly characterized in the case of the zeolite-alumina system,57 conceived to minimize the effects derived from the mismatching between the thermal expansions of its components. Although the ZIF-9-III material was experimentally revealed to have a higher average TEC than that predicted for ZIFs,28 both materials forming the membrane (ZIF and Ni) displayed moderate positive TECs. This therefore enhanced their compatibility and helped to inhibit the damage caused by the contraction/expansion phenomena through the temperature oscillations. This is a critical issue for membranes supported on polymeric HFs with even higher TECs, where changes in the glassy behavior of the polymer in terms of stiffness and brittleness must be taken into account.66 In this respect, Wang et al. have recently reported a strategy for ZIF-8 membrane fabrication on a Al2O3/ZnO HF support.67 The MOF-support adherence was improved because of the ZnO acting as nucleus and anchor for ZIF-8 growth. A H2/CO2 selectivity of 5.7 together with a high H2 permeance was obtained in this case. The use of ZnO particles to achieve a good adhesion of ZIF-8 layers with the support has also been made by Meckler et al.68 and Hess et al.69

4. Conclusions The effects of temperature on cobalt-benzimidazolate ZIF-9-III, a MOF material with promising properties for H2/CO2 separation, were studied by means of its insertion in a nickel hollow fiber resulting in a pore-plugging membrane. A significant increase in the unit cell volume in the -116 – 250 ºC range was inferred from XRD patterns for ZIF-9-III, as high as 5.5 %, and verified for the analogous material ZIF-7-III. This phenomenon can be linked with the experimental results obtained with the ZIF-9@Ni HF membrane in the H2/CO2 mixture separation: the contraction of the structure at low temperatures led to a H2/CO2 separation selectivity of 22.2 at -10 ºC. Therefore

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

a 3-fold higher apparent activation energy was calculated for CO2 than for H2. The TECs along each axis and for different temperature ranges were calculated for ZIF-7-III and ZIF-9-III, the first experimental values reported for a ZIF material. They showed large anisotropic shifts in the unit cell lengths, related with the structures of the studied materials, composed of stacked sheets. The matching between the signs of the overall TECs calculated for ZIF-9-III and that of the metal nickel in the support (i.e. both expand with temperature) would give the membrane a high degree of stability, durability and robustness, as demonstrated in the ZIF-9-III@Ni HF fabricated by microfluidics. Moreover, the 610 µm-OD metallic hollow fiber support and the resultant ZIF membrane simultaneously produced a higher separation selectivity and a better degree of energy and cost efficiency than those of inorganic-supported membranes described in the literature. Consequently, they are promising candidates for up-scaling.

5. Acknowledgements Financial support (MAT2013-40556-R, MAT2016-77290-R) from the Spanish MINECO and FEDER, the Aragón Government (DGA, T05) and the European Social Fund is gratefully acknowledged. F.C.-B. acknowledges his DGA predoctoral fellowship. We also acknowledge the use of the Servicio General de Apoyo a la Investigación-SAI (Universidad de Zaragoza). All the microscopy work was done in the Laboratorio de Microscopías Avanzadas at the Instituto de Nanociencia de Aragón (LMA-INA). The authors acknowledge the LMA-INA for offering access to their instruments and expertise.

6. References (1) Sorribas, S.; Gorgojo, P.; Téllez, C.; Coronas, J.; Livingston, A. G. High Flux Thin Film Nanocomposite Membranes Based on Metal–Organic Frameworks for Organic Solvent Nanofiltration. J. Am. Chem. Soc. 2013, 135 (40), 15201-15208.

22

ACS Paragon Plus Environment

Page 22 of 30

Page 23 of 30

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

(2) Qiao, X.; Chung, T. S.; Pramoda, K. P. Fabrication and Characterization of BTDA-TDI/MDI (P84) Co-Polyimide Membranes for the Pervaporation Dehydration of Isopropanol. J. Membr. Sci. 2005, 264 (1–2), 176-189. (3) Liu, R. X.; Qiao, X. Y.; Chung, T. S. Dual-Layer P84/Polyethersulfone Hollow Fibers for Pervaporation Dehydration of Isopropanol. J. Membr. Sci. 2007, 294 (1–2), 103-114. (4) Van Der Bruggen, B.; Vandecasteele, C.; Van Gestel, T.; Doyen, W.; Leysen, R. A Review of Pressure-Driven Membrane Processes in Wastewater Treatment and Drinking Water Production. Environ. Prog. 2003, 22 (1), 46-56. (5) Yampolskii, Y. Polymeric Gas Separation Membranes. Macromolecules 2012, 45 (8), 32983311. (6) Li, P.; Wang, Z.; Qiao, Z.; Liu, Y.; Cao, X.; Li, W.; Wang, J.; Wang, S. Recent Developments in Membranes for Efficient Hydrogen Purification. J. Membr. Sci. 2015, 495, 130-168. (7) Vanherck, K.; Koeckelberghs, G.; Vankelecom, I. F. J. Crosslinking Polyimides for Membrane Applications: A Review. Prog. Polym. Sci. 2013, 38 (6), 874-896. (8) Qiu, W.; Zhang, K.; Li, F. S.; Koros, W. J. Gas Separation Performance of Carbon Molecular Sieve Membranes Based on 6FDA-MPDA/DABA (3:2) Polyimide. ChemSusChem 2014, 7 (4), 1186-1194. (9) Nenoff, T. M. Hydrogen Purification: MOF Membranes Put to the Test. Nat. Chem. 2015, 7 (5), 377-378. (10) Adatoz, E.; Avci, A. K.; Keskin, S. Opportunities and Challenges of MOF-Based Membranes in Gas Separations. Sep. Purif. Technol. 2015, 152, 207-237. (11) Venna, S. R.; Carreon, M. A. Metal Organic Framework Membranes for Carbon Dioxide Separation. Chem. Eng. Sci. 2015, 124, 3-19. (12) Banerjee, R.; Furukawa, H.; Britt, D.; Knobler, C.; O’Keeffe, M.; Yaghi, O. M. Control of Pore Size and Functionality in Isoreticular Zeolitic Imidazolate Frameworks and Their Carbon Dioxide Selective Capture Properties. J. Am. Chem. Soc. 2009, 131 (11), 3875-3877.

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

(13) Banerjee, R.; Phan, A.; Wang, B.; Knobler, C.; Furukawa, H.; O'Keeffe, M.; Yaghi, O. M. High-Throughput Synthesis of Zeolitic Imidazolate Frameworks and Application to CO2 Capture. Science 2008, 319 (5865), 939-943. (14) Zornoza, B.; Tellez, C.; Coronas, J.; Gascon, J.; Kapteijn, F. Metal Organic Framework Based Mixed Matrix Membranes: An Increasingly Important Field of Research with a Large Application Potential. Microporous Mesoporous Mater. 2013, 166, 67-78. (15) Li, W.; Su, P.; Zhang, G.; Shen, C.; Meng, Q. Preparation of Continuous NH2-MIL-53 Membrane on Ammoniated Polyvinylidene Fluoride Hollow Fiber for Efficient H2 Purification. J. Membr. Sci. 2015, 495, 384-391. (16) Kwon, H. T.; Jeong, H. K.; Lee, A. S.; An, H. S.; Lee, J. S. Heteroepitaxially Grown Zeolitic Imidazolate Framework Membranes with Unprecedented Propylene/Propane Separation Performances. J. Am. Chem. Soc. 2015, 137 (38), 12304-12311. (17) Cacho-Bailo, F.; Caro, G.; Etxeberría-Benavides, M.; Karvan, O.; Tellez, C.; Coronas, J. High Selectivity ZIF-93 Hollow Fiber Membranes for Gas Separation. Chem. Commun. 2015, 51, 11283-11285. (18) Zhang, C.; Koros, W. J. Zeolitic Imidazolate Framework-Enabled Membranes: Challenges and Opportunities. J. Phys. Chem. Lett. 2015, 6 (19), 3841-3849. (19) Robeson, L. M. The Upper Bound Revisited. J. Membr. Sci. 2008, 320 (1–2), 390-400. (20) Brown, A. J.; Brunelli, N. A.; Eum, K.; Rashidi, F.; Johnson, J. R.; Koros, W. J.; Jones, C. W.; Nair, S. Interfacial Microfluidic Processing of Metal-Organic Framework Hollow Fiber Membranes. Science 2014, 345 (6192), 72-75. (21) Cacho-Bailo, F.; Catalán-Aguirre, S.; Etxeberría-Benavides, M.; Karvan, O.; Sebastian, V.; Téllez, C.; Coronas, J. Metal-Organic Framework Membranes on the Inner-Side of a Polymeric Hollow Fiber by Microfluidic Synthesis. J. Membr. Sci. 2015, 476, 277-285. (22) Lin, R.; Ge, L.; Hou, L.; Strounina, E.; Rudolph, V.; Zhu, Z. Mixed Matrix Membranes with Strengthened MOFs/Polymer Interfacial Interaction and Improved Membrane Performance. ACS Appl. Mater. Interfaces 2014, 6 (8), 5609-5618.

24

ACS Paragon Plus Environment

Page 24 of 30

Page 25 of 30

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

(23) Semino, R.; Ramsahye, N. A.; Ghoufi, A.; Maurin, G. Microscopic Model of the MetalOrganic Framework/Polymer Interface: A First Step toward Understanding the Compatibility in Mixed Matrix Membranes. ACS Appl. Mater. Interfaces 2016, 8 (1), 809-819. (24) Cacho-Bailo, F.; Tellez, C.; Coronas, J. Interactive Thermal Effects on Metal-Organic Framework Polymer Composite Membranes. Chem. Eur. J. 2016, 22 (28), 9533-9536. (25) David, O.; Gendel, Y.; Wessling, M. Tubular Macro-Porous Titanium Membranes. J. Membr. Sci. 2014, 461, 139-145. (26) Meng, B.; Tan, X.; Meng, X.; Qiao, S.; Liu, S. Porous and Dense Ni Hollow Fibre Membranes. J. Alloys Compd. 2009, 470 (1–2), 461-464. (27) Luiten-Olieman, M. W. J.; Winnubst, L.; Nijmeijer, A.; Wessling, M.; Benes, N. E. Porous Stainless Steel Hollow Fiber Membranes Via Dry–Wet Spinning. J. Membr. Sci. 2011, 370 (1– 2), 124-130. (28) Bouëssel du Bourg, L.; Ortiz, A. U.; Boutin, A.; Coudert, F.-X. Thermal and Mechanical Stability of Zeolitic Imidazolate Frameworks Polymorphs. APL Materials 2014, 2 (12), 124110. (29) Schneemann, A.; Bon, V.; Schwedler, I.; Senkovska, I.; Kaskel, S.; Fischer, R. A. Flexible Metal-Organic Frameworks. Chem. Soc. Rev. 2014, 43 (16), 6062-6096. (30) Zhou, W.; Wu, H.; Yildirim, T.; Simpson, J. R.; Walker, A. R. H. Origin of the Exceptional Negative Thermal Expansion in Metal-Organic Framework-5 Zn4O(1,4-Benzenedicarboxylate)3. Phys. Rev. B 2008, 78 (5), 054114. (31) Han, S. S.; Goddard, W. A. Metal−Organic Frameworks Provide Large Negative Thermal Expansion Behavior. J. Phys. Chem. C 2007, 111 (42), 15185-15191. (32) Woodcock, D. A.; Lightfoot, P.; Villaescusa, L. A.; Díaz-Cabañas, M.-J.; Camblor, M. A.; Engberg, D. Negative Thermal Expansion in the Siliceous Zeolites Chabazite and ITQ-4:  A Neutron Powder Diffraction Study. Chem. Mater. 1999, 11 (9), 2508-2514. (33) P. Attfield, M. Strong Negative Thermal Expansion in Siliceous Faujasite. Chem. Commun. 1998, (5), 601-602.

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

Page 26 of 30

(34) Marinkovic, B. A.; Jardim, P. M.; Saavedra, A.; Lau, L. Y.; Baehtz, C.; de Avillez, R. R.; Rizzo, F. Negative Thermal Expansion in Hydrated HZSM-5 Orthorhombic Zeolite. Microporous Mesoporous Mater. 2004, 71 (1–3), 117-124. (35) Bhange, D. S.; Ramaswamy, V. Negative Thermal Expansion in Silicalite-1 and Zirconium Silicalite-1 Having MFI Structure. Mater. Res. Bull. 2006, 41 (7), 1392-1402. (36) Noack, M.; Schneider, M.; Dittmar, A.; Georgi, G.; Caro, J. The Change of the Unit Cell Dimension of Different Zeolite Types by Heating and Its Influence on Supported Membrane Layers. Microporous Mesoporous Mater. 2009, 117 (1–2), 10-21. (37) Dong, J.; Lin, Y. S.; Hu, M. Z. C.; Peascoe, R. A.; Payzant, E. A. Template-RemovalAssociated

Microstructural

Development

of

Porous-Ceramic-Supported

MFI

Zeolite

Membranes. Microporous Mesoporous Mater. 2000, 34 (3), 241-253. (38) Cai, W.; Katrusiak, A. Giant Negative Linear Compression Positively Coupled to Massive Thermal Expansion in a Metal–Organic Framework. Nat. Commun. 2014, 5, 4337. (39) Adak, S.; Daemen, L. L.; Hartl, M.; Williams, D.; Summerhill, J.; Nakotte, H. Thermal Expansion in 3D-Metal Prussian Blue Analogs—a Survey Study. J. Solid State Chem. 2011, 184 (11), 2854-2861. (40) Ogborn, J. M.; Collings, I. E.; Moggach, S. A.; Thompson, A. L.; Goodwin, A. L. Supramolecular Mechanics in a Metal-Organic Framework. Chem. Sci. 2012, 3 (10), 3011-3017. (41) Baker, R. W. Future Directions of Membrane Gas Separation Technology. Ind. Eng. Chem. Res. 2002, 41 (6), 1393-1411. (42) Haszeldine, R. S. Carbon Capture and Storage: How Green Can Black Be? Science 2009, 325 (5948), 1647-1652. (43) Li, H.; Song, Z.; Zhang, X.; Huang, Y.; Li, S.; Mao, Y.; Ploehn, H. J.; Bao, Y.; Yu, M. Ultrathin, Molecular-Sieving Graphene Oxide Membranes for Selective Hydrogen Separation. Science 2013, 342 (6154), 95-98. (44) Sumida, K.; Rogow, D. L.; Mason, J. A.; McDonald, T. M.; Bloch, E. D.; Herm, Z. R.; Bae, T. H.; Long, J. R. Carbon Dioxide Capture in Metal–Organic Frameworks. Chem. Rev. 2012, 112 (2), 724-781.

26

ACS Paragon Plus Environment

Page 27 of 30

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

(45) Li, Y. S.; Liang, F. Y.; Bux, H.; Yang, W.; Caro, J. Zeolitic Imidazolate Framework ZIF-7 Based Molecular Sieve Membrane for Hydrogen Separation. J. Membr. Sci. 2010, 354 (1–2), 4854. (46) Cacho-Bailo, F.; Matito-Martos, I.; Perez-Carbajo, J.; Etxeberria-Benavides, M.; Karvan, O.; Sebastian, V.; Calero, S.; Tellez, C.; Coronas, J. On the Molecular Mechanisms for the H2/CO2 Separation Performance of Zeolite Imidazolate Framework Two-Layered Membranes. Chem. Sci. 2017, 8, 325-333. (47) Zhang, C.; Xiao, Y.; Liu, D.; Yang, Q.; Zhong, C. A Hybrid Zeolitic Imidazolate Framework Membrane by Mixed-Linker Synthesis for Efficient CO2 Capture. Chem. Commun. 2013, 49 (6), 600-602. (48) Zhao, P.; Lampronti, G. I.; Lloyd, G. O.; Wharmby, M. T.; Facq, S.; Cheetham, A. K.; Redfern, S. A. T. Phase Transitions in Zeolitic Imidazolate Framework 7: The Importance of Framework Flexibility and Guest-Induced Instability. Chem. Mater. 2014, 26 (5), 1767-1769. (49) Yang, Q.-F.; Cui, X.-B.; Yu, J.-H.; Lu, J.; Yu, X.-Y.; Zhang, X.; Xu, J.-Q.; Hou, Q.; Wang, T.-G. A Series of Metal-Organic Complexes Constructed from in Situ Generated Organic Amines. CrystEngComm 2008, 10 (11), 1534-1541. (50) He, M.; Yao, J.; Li, L.; Wang, K.; Chen, F.; Wang, H. Synthesis of Zeolitic Imidazolate Framework-7 in a Water/Ethanol Mixture and its Ethanol-Induced Reversible Phase Transition. ChemPlusChem 2013, 78 (10), 1222-1225. (51) Peng, Y.; Li, Y.; Ban, Y.; Jin, H.; Jiao, W.; Liu, X.; Yang, W. Metal-Organic Framework Nanosheets as Building Blocks for Molecular Sieving Membranes. Science 2014, 346 (6215), 1356-1359. (52) Cacho-Bailo, F.; Caro, G.; Etxeberria, M.; Karvan, O.; Tellez, C.; Coronas, J. MOF-Polymer Enhanced Compatibility: Post-Annealed Zeolite Imidazolate Framework Membranes inside Polyimide Hollow Fibers. RSC Adv. 2016, 6, 5881-5889. (53) Diamond - Crystal and Molecular Structure Visualization; Crystal Impact - Dr. H. Putz & Dr. K. Brandenburg GbR, K., 53227 Bonn, Germany; http://www.crystalimpact.com/diamond.

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

(54) Sorenson, S. G.; Smyth, J. R.; Kocirik, M.; Zikanova, A.; Noble, R. D.; Falconer, J. L. Adsorbate-Induced Expansion of Silicalite-1 Crystals. Ind. Eng. Chem. Res. 2008, 47 (23), 96119616. (55) Tan, J.-C.; Civalleri, B.; Lin, C.-C.; Valenzano, L.; Galvelis, R.; Chen, P.-F.; Bennett, T. D.; Mellot-Draznieks, C.; Zicovich-Wilson, C. M.; Cheetham, A. K. Exceptionally Low Shear Modulus in a Prototypical Imidazole-Based Metal-Organic Framework. Phys. Rev. Lett. 2012, 108 (9), 095502. (56) Wang, S.; Hou, Y.; Lin, S.; Wang, X. Water Oxidation Electrocatalysis by a Zeolitic Imidazolate Framework. Nanoscale 2014, 6 (17), 9930-9934. (57) Miachon, S.; Landrivon, E.; Aouine, M.; Sun, Y.; Kumakiri, I.; Li, Y.; Prokopová, O. P.; Guilhaume, N.; Giroir-Fendler, A.; Mozzanega, H.; Dalmon, J. A. Nanocomposite MFI-Alumina Membranes Via Pore-Plugging Synthesis: Preparation and Morphological Characterisation. J. Membr. Sci. 2006, 281 (1–2), 228-238. (58) Sarkisov, L.; Martin, R. L.; Haranczyk, M.; Smit, B. On the Flexibility of Metal–Organic Frameworks. J. Am. Chem. Soc. 2014, 136 (6), 2228-2231. (59) Fairen-Jimenez, D.; Moggach, S. A.; Wharmby, M. T.; Wright, P. A.; Parsons, S.; Düren, T. Opening the Gate: Framework Flexibility in ZIF-8 Explored by Experiments and Simulations. J. Am. Chem. Soc. 2011, 133 (23), 8900-8902. (60) Mao, Y.; Li, J.; Cao, W.; Ying, Y.; Sun, L.; Peng, X. Pressure-Assisted Synthesis of HKUST1 Thin Film on Polymer Hollow Fiber at Room Temperature toward Gas Separation. ACS Appl. Mater. Interfaces 2014, 6 (6), 4473-4479. (61) Firpo, G.; Angeli, E.; Repetto, L.; Valbusa, U. Permeability Thickness Dependence of Polydimethylsiloxane (PDMS) Membranes. J. Membr. Sci. 2015, 481, 1-8. (62) Sandström, L.; Sjöberg, E.; Hedlund, J. Very High Flux MFI Membrane for CO2 Separation. J. Membr. Sci. 2011, 380 (1–2), 232-240. (63) Aguado, S.; Bergeret, G.; Titus, M. P.; Moizan, V.; Nieto-Draghi, C.; Bats, N.; Farrusseng, D. Guest-Induced Gate-Opening of a Zeolite Imidazolate Framework. New J. Chem. 2011, 35 (3), 546-550.

28

ACS Paragon Plus Environment

Page 28 of 30

Page 29 of 30

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

(64) Aguado, S.; Canivet, J.; Farrusseng, D. Facile Shaping of an Imidazolate-Based MOF on Ceramic Beads for Adsorption and Catalytic Applications. Chem. Commun. 2010, 46 (42), 79998001. (65) Wang, N.; Liu, Y.; Qiao, Z.; Diestel, L.; Zhou, J.; Huang, A.; Caro, J. Polydopamine-Based Synthesis of a Zeolite Imidazolate Framework ZIF-100 Membrane with High H2/CO2 Selectivity. J. Mater. Chem. A 2015, 3 (8), 4722-4728. (66) Mahieux, C. A.; Reifsnider, K. L. Property Modeling across Transition Temperatures in Polymers: A Robust Stiffness–Temperature Model. Polymer 2001, 42 (7), 3281-3291. (67) Wang, X.; Sun, M.; Meng, B.; Tan, X.; Liu, J.; Wang, S.; Liu, S. Formation of Continuous and Highly Permeable ZIF-8 Membranes on Porous Alumina and Zinc Oxide Hollow Fibers. Chem. Commun. 2016, 52 (92), 13448-13451. (68) Meckler, S. M.; Li, C.; Queen, W. L.; Williams, T. E.; Long, J. R.; Buonsanti, R.; Milliron, D. J.; Helms, B. A. Sub-Micron Polymer–Zeolitic Imidazolate Framework Layered Hybrids Via Controlled Chemical Transformation of Naked ZnO Nanocrystal Films. Chem. Mater. 2015, 27 (22), 7673-7679. (69) Hess, S. C.; Grass, R. N.; Stark, W. J. MOF Channels within Porous Polymer Film: Flexible, Self-Supporting ZIF-8 Poly(Ether Sulfone) Composite Membrane. Chem. Mater. 2016, 28 (21), 7638-7644.

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 Table of Contents only

30

ACS Paragon Plus Environment

Page 30 of 30