Letter pubs.acs.org/NanoLett
Cite This: Nano Lett. XXXX, XXX, XXX-XXX
Dual-Channel, Molecular-Sieving Core/Shell ZIF@MOF Architectures as Engineered Fillers in Hybrid Membranes for Highly Selective CO2 Separation Zhuonan Song,†,∥ Fen Qiu,†,∥ Edmond W. Zaia,† Zhongying Wang,‡ Martin Kunz,§ Jinghua Guo,§ Michael Brady,§ Baoxia Mi,‡ and Jeffrey J. Urban*,† †
The Molecular Foundry, Lawrence Berkeley National Laboratory, Berkeley, California 94720, United States Civil and Environmental Engineering, University of California, Berkeley, California 94720, United States § Advanced Light Source, Lawrence Berkeley National Laboratory, Berkeley, California 94720, United States ‡
S Supporting Information *
ABSTRACT: A novel core/shell porous crystalline structure was prepared using a large pore metal organic framework (MOF, UiO-66-NH2, pore size, ∼ 0.6 nm) as core surrounded by a small pore zeolitic imidazolate framework (ZIF, ZIF-8, pore size, ∼ 0.4 nm) through a layer-by-layer deposition method and subsequently used as an engineered filler to construct hybrid polysulfone (PSF) membranes for CO2 capture. Compared to traditional fillers utilizing only one type of porous material with rigid channels (either large or small), our custom designed core/shell fillers possess clear advantages via pore engineering: the large internal channels of the UiO-66-NH2 MOFs create molecular highways to accelerate molecular transport through the membrane, while the thin shells with small pores (ZIF-8) or even smaller pores generated at the interface by the imperfect registry between the overlapping pores of ZIF and MOF enhance molecular sieving thus serving to distinguish slightly larger N2 molecules (kinetic diameter, 0.364 nm) from smaller CO2 molecules (kinetic diameter, 0.33 nm). The resultant core/shell ZIF@ MOF and as-prepared hybrid PSF membranes were characterized by transmission electron microscopy, X-ray diffraction, wideangle X-ray scattering, scanning electron microscopy, Fourier transform infrared, thermogravimetric analysis, differential scanning calorimetry, and contact angle tests. The dependence of the separation performance of the membranes on the MOF/ZIF ratio was also studied by varying the number of layers of ZIF coatings. The integrated PSF-ZIF@MOF hybrid membrane (40 wt % loading) with optimized ZIF coating cycles showed improved hydrophobicity and excellent CO2 separation performance by simultaneously increasing CO2 permeability (CO2 permeability of 45.2 barrer, 710% higher than PSF membrane) and CO2/N2 selectivity (CO2/N2 selectivity of 39, 50% higher than PSF membrane), which is superior to most reported hybrid PSF membranes. The strategy of using dual-channel molecular sieving core/shell porous crystals in hybrid membranes thus provides a promising means for CO2 capture from flue gas. KEYWORDS: Metal organic frameworks, core/shell, hybrid membrane, CO2 capture
T
composition: 70−75% N2, 15−16% CO2, 5−7% H2O, and 3− 4% O2) from existing power plants accounts for roughly onethird of total CO2 emissions.5,6 To reduce worldwide CO2 accumulation meaningfully, postcombustion CO2 capture from flue gas (predominantly CO2/N2 separation) has been
he concentration of CO2 in the atmosphere has increased by almost 50% since the start of the Industrial Revolution.1 Mitigating the sharply rising level of atmospheric CO2, directly correlated to the global warming and ocean acidification, remains an ongoing environmental concern facing our civilization.2,3 However, the rapidly increasing trend of CO2 emissions resulting from our heavy use of fossil fuels may not be alleviated in the foreseeable future due to global energy demand, particularly in developing nations.4 Flue gas (volume © XXXX American Chemical Society
Received: July 9, 2017 Revised: October 22, 2017 Published: October 26, 2017 A
DOI: 10.1021/acs.nanolett.7b02910 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 1. Schematic illustration of core/shell ZIF@MOF, hybrid PSF membrane preparation, and optical image of a 40 wt % PSF-5-ZIF@MOF membrane.
solids to produce hybrid membranes.21−24 Over approximately the past decade, porous crystals such as metal organic frameworks (MOFs) and zeolitic imidazolate frameworks (ZIFs), have received growing interest as potential fillers in hybrid membranes. MOFs consist of metal cations or clusters linked by organic linkers forming a porous crystalline network.25 ZIFs, a subclass of MOFs, consist of transition metal cations (mainly tetrahedral Zn2+) bridged by imidazolebased ligands.26 One strong advantage of using MOFs as the inorganic fillers is that MOFs can possess good compatibility with polymer chains due to their partially organic nature, resulting in hybrid membranes with better interfacial wetting.27 In hybrid membrane design, it is crucial to select a filler with appropriate pore size for the relevant separation of interest.28 So far, more than 20,000 types of MOFs have been created with pore sizes ranging from 0.3 to over 10 nm.29 However, despite a large selection of MOFs to choose from there are not often MOFs with ideal pore sizes for many of the most crucial separation challenges today. For example, current MOF technology is not ideal for separating molecules that are very similar in kinetic diameter, like CO2 (kinetic diameter, 0.33)30 from N2 (kinetic diameter, 0.364 nm);30 these two molecules differ in kinetic radius by less than 10%. Large-pore MOFs can facilitate rapid diffusion of gas molecules, while smaller pore MOFs may help realize better size-sieving separations but hinder gas transport. For example, the large-pore (>0.5 nm) MOF, UiO-66-NH2 (∼0.6 nm), has been introduced in hybrid membranes by our group recently to increase CO2 permeability but this was at the sacrifice of some CO2/N2 selectivity.16 Likewise, hybrid membranes containing small pore ZIFs (ZIF90, ZIF-8) have been studied elsewhere and display increased CO2 selectivity over other gases, at the expense of reduced permeability.31−33 Therefore, a new material paradigm is necessary to address this permeability/selectivity trade-off. Herein, we propose a new architecture for porous crystals to solve separation challenges in hybrid membranes−core/shell MOF systems. Instead of utilizing only one type of MOF as a filler, we report the first successful synthesis and implementation of two types of MOFs into a novel core/shell structural
identified as the greatest near-term potential route and can also be readily retrofitted to existing state-of-the-art power plants.2,7−10 Many types of aqueous alkanolamine solutions (amine scrubbers) have been studied and employed in conventional CO2 removal via absorption processes. These liquid-based processes capture CO2 with a high selectivity but suffer in their regeneration step due to the high energy penalty to reverse the CO2-amine binding reaction.10,11 Moreover, the corrosive and volatile properties of amine solutions may result in additional pollutant emissions.12,13 The regeneration energy required for CO2 capture by physical adsorption using porous materials can be significantly reduced compared to aqueous amine-based absorption processes due to the intrinsically lower heat capacities of solids than water.14 However, these solid sorbents are typically prepared as powders that can be difficult to implement CO2 capture at a viable industrial scale due to the high pressure drops and poor heat transfer issues.15 Relative to the above-mentioned conventional approaches, membrane-based gas separation processes have attracted intense interest because of their superior cost-effectiveness, environmental friendliness, and simplicity for postcombustion CO2 capture.16 Historically, polymers have been the dominant commercial materials used in gas separations due to their ease of fabrication and scalable production, and relatively low cost.17 Unfortunately, there is a traditional trade-off between permeability and selectivity for polymer membranes: highly selective membranes tend to have low permeability and vice versa. Inorganic membranes, on the other hand, have the potential to offer a better balance between permeability and selectivity.18 However, they suffer from problems of brittleness, high cost, and complex fabrication processes that still require further optimization.19 In order to overcome these limitations, hybrid membranes, or so-called mixed matrix membranes, have been proposed as a promising solution; these hybrid systems combine the advantages of both polymeric and inorganic membranes into one system.20 Indeed, various polymers have been modified with inorganic fillers such as zeolites, mesoporous silicas, carbon nanotubes, and even nonporous B
DOI: 10.1021/acs.nanolett.7b02910 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters filler for hybrid membrane preparation, in which a large-pore MOF (UiO-66-NH2) acts as the core to provide facilitated transport pathways; small-pore ZIF (ZIF-8, effective pore aperture of ∼0.4 nm)34 or even smaller pores generated at the interface by incommensurate overlapping of the pores of MOF and ZIF, which will act as a molecular filter to enhance molecular sieving properties. An experimental analogy exists in the zeolite field; it has been proven recently that the reduction of zeolite pore mouths can be achieved by depositing a layer of microporous coating on the external surface.35,36 Similar strategy has also been applied on membrane and catalyst modification to narrow the effective pore size and increase selectivity.37,38 Moreover, the hydrophobic nature of ZIF-8 may provide an additional benefit at the macroscopic level, by more efficiently reducing the uptake of water vapor from the flue gas. The preparation of UiO-66-NH2 (MOF) was carried out by following a microwave synthetic technique (Supporting Information). Zinc nitrate hexahydrate (Zn(NO3)2·6H2O, >98%) and 2-methylimidazole (>99%) were purchased from Sigma-Aldrich. Methanol (99.8%) was purchased from BDH Chemicals. The ZIF-8 coating was prepared on the external surface of UiO-66-NH2 using a layer-by-layer solution deposition method at room temperature by introducing the UiO-66-NH2 into the inorganic and organic precursors of ZIF8 sequentially (and separately) as shown in Figure 1. The amino group on the UiO-66-NH2 enables this layer-by-layer processing by acting as a covalent linker between the MOF and ZIF precursors. 39,40 A layer-by-layer method of ZIF-8 preparation was already demonstrated by Eddaoudi on Au substrates, but never before on MOFs.41 Typically, 0.25 g of UiO-66-NH2 powder was dispersed in an inorganic precursors solution of ZIF (0.219 g of Zn(NO3)2·6H2O dissolved in 10 mL of methanol) with stirring for 2 h and then washed with methanol. After centrifugation, the inorganic precursorenriched MOF powder was then immersed in the organic precursor solution of ZIF (0.486 g of 2-methylimidazole dissolved in 10 mL of methanol) with stirring for another 2 h and then washed with solvent afterward. The ZIF-coated MOF powder was then moved into a 75 °C oven to evaporate the solvent and form ZIF. This process was carried out once for 1ZIF@MOF and then cycled 3, 5, and 8 times for 3-ZIF@MOF, 5-ZIF@MOF and 8-ZIF@MOF, respectively. Pristine ZIF-8 samples were also synthesized for comparison (Supporting Information). Before membrane preparation, polysulfone (Udel P-1700, kindly supplied by Solvay Plastics) was degassed at 110 °C for 12 h under vacuum to remove adsorbed water. For the pure polymer membrane, 0.366 g of polysulfone (PSF) was dissolved in 4.0 mL of chloroform (BDH Chemicals) and stirred for 24 h to form a viscous solution and subsequently filtered with a 0.2 μm PVDF filter (Pall). The fabrication procedure for the hybrid membranes was identical to that of the pure PSF membrane but with an additional dispersion stage. First, the as-synthesized MOF or ZIF@MOF was dispersed in chloroform by a tip sonicator (Sonics & Materials). Once dispersed, one-fifth of the PSF solution was added subsequently during sonication to increase interaction and homogeneity between porous fillers and polymer by coating the fillers with a thin polymer layer. The remaining PSF solution was then added afterward and sonicated. The solid concentration of final solution was adjusted to around 30 wt % by evaporation of excess solvent to minimize sedimentation. After sonication to remove the air bubbles, the resulting PSF-MOF (hybrid PSF
membranes containing pure MOF) or PSF-ZIF@MOF (hybrid PSF membranes containing core/shell ZIF@MOF) solution was cast into a casting plate (Figure 1), loosely covered, and allowed to dry under atmospheric conditions over 48 h. The dried membranes are then placed into a vacuum oven at 110 °C overnight to remove any residual solvent and water. Hybrid membranes were prepared with fixed loading of 40 wt % ZIF@ MOF but with different ZIF coating cycles. For instance, a defect-free, 7 cm in diameter, hybrid membrane containing 40 wt % of 5 cycles ZIF coated MOF (PSF-5-ZIF@MOF, abbreviation will be used in the following description) is shown in Figure 1. The thickness of the final pure and hybrid membranes varied within the range of 60−70 um as measured using a micrometer (Mitutoyo Corp). Morphologies of the as-prepared UiO-66-NH2 (MOF) and core/shell ZIF@MOF nanoparticles were examined by transmission electron microscopy (TEM) and scanning electron microscopy (SEM) together with energy dispersive X-ray spectroscopy (EDS). The crystalline structures of MOF and ZIF@MOF were measured by powder X-ray diffraction (PXRD) and transmission wide-angle X-ray scattering (WAXS). The functional groups of synthesized MOF, ZIF, and ZIF@MOF were further characterized by attenuated total reflection (ATR, PerkinElmer Spectrum One) spectra within the range of 600−4000 cm−1. The thermal properties of MOF, ZIF and ZIF@MOF were characterized by Q5000 thermogravimetric analysis (TGA) in the range of room temperature to 600 °C with the rate of 10 °C min−1. Adsorption experiments of MOF, ZIF, and ZIF@MOF were carried out by Micromeritics ASAP2020 with CO2 and Tristar II (Micromeritics) with N2 at 25 °C up to a pressure of 1 bar. Before the measurements, the sample was degassed under vacuum at 110 °C for 12 h. Differential scanning calorimetry (DSC, Q2000, TA Instruments, U.S.A.) measurements were performed from 40 to 250 °C in N2 atmosphere to study the glass transition temperatures (Tg) of the membranes. The water contact angle (CA) was measured at room temperature using Attension Theta Lite (Biolin Scientific) to determine hydrophilic and hydrophobic characteristics of the prepared membranes. Density measurements of the hybrid membranes are conducted with a density determination kit (Mettler Toledo) following Archimedes’ principle (Supporting Information). Pure gas permeabilities (CO2 and N2) of PSF, PSF-MOF, and PSF-ZIF@MOF membranes were tested using a variable pressure constant volume gas permeation apparatus at 35 °C. The membranes are masked with brass discs to accurately define an area where gas transport can occur. Prior to testing, the membranes are degassed within the apparatus overnight. A fixed pressure is applied to the upstream side of the membrane, while the gas flux is recorded as a steady-state pressure rise downstream of the membrane. The gas permeability was calculated according to P=
VDl ⎛ dp1 ⎞ ⎟ ⎜ p2 ART ⎝ dt ⎠
where P is the gas permeability [1 Barrer = 10−10 cm3 (STP) cm/(cm2·s·cmHg)], VD represents the downstream volume (cm3), l refers to the film thickness (cm), p2 indicates the upstream pressure (cmHg), A is the effective membrane area (cm2), R is the gas constant, T is the operating temperature (K), and dp1/dt is steady state pressure rise downstream at fixed upstream pressure (cmHg/s). The ideal selectivity of C
DOI: 10.1021/acs.nanolett.7b02910 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 2. TEM image of MOF (a) and 5-ZIF@MOF (b); SEM image of 5-ZIF@MOF (the inset image shows EDS spot scan spectrum) (c); EDS line scan along the 5-ZIF@MOF particle (d); ATR spectra of UiO-66-NH2, ZIF-8, and ZIF@MOF (e); and CO2 and N2 adsorption isotherms at 25 °C for UiO-66-NH2, ZIF-8, and ZIF@MOF (f).
S3). The resulting peaks corresponding to the (200) plane at 2theta = 10.6 and (112) plane at 2-theta = 12.7 were detected on the 8-ZIF@MOF sample by XRD and WAXS, proving that the ZIF shell crystal was successfully grown on UiO-66-NH2 (Figure S3). Those two representative peaks only emerge in the 8-ZIF@MOF samples, which is due to the fact that the ZIF crystalline peaks require a certain amount of shelling thickness to forming strongly scattering crystalline peaks in this core/ shell structure. ATR analysis of pure UiO-66-NH2, pure ZIF-8, and ZIF@ MOF with different ZIF coating cycles were performed as shown in Figure 2e. The spectrum of pure UiO-66-NH2 matches well with the pattern reported in the literature.42 The peak at around 1620 cm−1 corresponds to the N−H bending vibration.43 The intense doublet at 1415 and 1387 cm−1 can be assigned to the stretching modes of the carboxylic groups.44 For pure ZIF-8, the peak at 1590 cm−1 is attributed to the CN stretching.45 In the lower frequency region, an additional peak at around 1150 cm−1 is observed from ATR spectra for ZIF-8 and ZIF@MOF samples, which is associated with the in-plane bending of imidazole ring in the ZIF-8.45 Moreover, the intensity of this characteristic peak grows stronger with the increase of ZIF cycle numbers, further confirming the successful growth of ZIF shell around MOF. Figure 2f shows the experimental CO2 and N2 adsorption isotherms at 25 °C of pure UiO-66-NH2, ZIF-8, and ZIF@ MOF. UiO-66-NH2 showed the highest CO2 adsorption uptake and ZIF-8 showed the lowest, in good agreement with previous reports.46,47 Therefore, the CO2 adsorption uptake of the core/ shell ZIF@MOF decreased gradually with more ZIF coating cycles. However, after five cycles of ZIF coating, the 5-ZIF@ MOF still retained good CO2 adsorption capacity (2.33 mmol/
CO2/N2 can be calculated by the ratio of the permeability of the individual gas. Figure 2a,b shows the high-resolution TEM (Supporting Information) images of MOF and core/shell ZIF@MOF. It was observed that, after 5 cycles of ZIF coating, an ultrathin (∼20 nm) layer of ZIF shell was continuously grown on the MOF surface. High-resolution SEM images (Supporting Information) showed that the synthesized MOF and 5-ZIF@MOF are mainly cubic crystals (Figure S1). The smooth surface of the MOF (Figure S1a) becomes much rougher after five cycles of ZIF coating (Figure S1b), also suggesting the coverage of ZIF-8 shell on the surface of MOF particles. Both EDS spot scans (labeled as #1) and line scans (labeled as #2) were carried out using Phenom ProX (SEM) to analyze the elemental composition of synthesized MOF (Figure S2) and 5-ZIF@ MOF samples (Figure 2c,d). It can be seen that the pure MOF is composed of Zr, C, and O elements, which is consistent with its structure (Figure S2). The observed Si comes from the substrate. For 5-ZIF@MOF, a new peak of Zn element appears at around 1 keV from the spot scan as shown in the inset of Figure 2c, indicating the successful deposition of ZIF coating on the surface of MOF. Figure 2d shows the EDS line scan along the 5-ZIF@MOF particles. The clear Zn signal when compared to the line scan displaced for pure MOF (Figure S2b) further demonstrates the core/shell structure of ZIF@ MOF. In order to further prove that the crystalline structure of ZIF-8 shell has be successfully grown on the MOF core, both PXRD and transmission WAXS of the five samples (UiO-66NH2, 1-ZIF@MOF, 3-ZIF@MOF, 5-ZIF@MOF, and 8-ZIF@ MOF) have been measured. The results of XRD and WAXS on a Cu K-α scale match the crystalline patterns of both UiO-66NH2 and ZIF-8 in a range of 2θ value from 5° to 50° (Figure D
DOI: 10.1021/acs.nanolett.7b02910 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
shift of Tg by ∼5 °C (Figure S5). This positive shift is due to reduced polymer chain mobility and rigidification as the polymer is adsorbed onto the filler surface, indicating favorable interactions and improved mechanical properties.48 The density of the membrane was calculated and shown in Figure S6. A linear density decrease trend was observed for hybrid membranes as ZIF coating cycles increased due to the smaller density of ZIF-8 (0.95 g/cm3).49 The improvements of surface hydrophobicity of prepared membranes were investigated by contact angle (CA) measurements. As shown in Figure 3g,h for neat PSF and PSF-MOF membranes, CAs were measured to be 78.5 ± 0.5° and 83.5 ± 0.7° respectively, revealing a hydrophilic property of the surface. However, for PSF-5-ZIF@MOF hybrid membrane, the CA increased to around 96° (Figure 3i), converting the surface from hydrophilic to hydrophobic, due to the good coverage of ZIF-8 shell. Consequently, we expect the PSF-ZIF@MOF hybrid membranes would more efficiently hinder the entrance of water molecules into the membrane in real flue gas capture. Pure gas permeability and selectivity at 35 °C and 3 bar are shown in Figure 4. The neat PSF membrane shows a CO2 permeability of 5.6 barrer and CO2/N2 selectivity of 26. After incorporating 40 wt % of MOF (UiO-66-NH2), the CO2 permeability dramatically increased to 48.2 barrer or 760% higher permeability than the permeability of neat PSF membrane, which is consistent with formation of a percolative network of MOF crystals, as previously reported.16 However, the CO2/N2 selectivity of PSF-MOF hybrid membrane decreased to 25, which may result from the poor molecular sieving effect of UiO-66-NH2 (0.6 nm pore size) for distinguishing slightly bigger N2 molecules over CO2. When we applied the external ZIF coating (as shown in Figure 1), after the first three cycles (PSF-3-ZIF@MOF) the CO2/N2 selectivity increased to 28 due to the slightly decreased N2 permeability. When the ZIF coating cycles increased to 5, the N2 permeability decreased even faster, while maintaining a relatively high CO2 permeability of 45.2 barrer (710% higher than PSF). This resulted in a sharp increase in selectivity to 39 (CO2/N2), which is 50% higher than the neat PSF membrane and 56% higher than PSF-MOF hybrid membranes. As we further increased the ZIF coating to eight cycles, the high CO2/ N2 selectivity was maintained. However, a large decrease of CO2 permeability was observed. The permeation results suggest that with five cycles of ZIF coating, a thin layer of ZIF-8 forms on the external surface of MOF and increases the molecular sieving effect for CO2 over N2. After five cycles, as we further increased the number of ZIF coating layers, CO2 permeation goes down due to the increased transport resistance. Coupled with the large increases in both gas permeability and selectivity, the separation performance of hybrid PSF membranes containing ZIF@MOF as fillers is significantly enhanced over neat PSF membrane and moves much closer toward the Robeson upper bound line (Figure S7). So far, most of the state-of-the-art selective membranes for CO2/N2 separation exceeding the Robeson upper bound are inorganic membranes, such as NaY, ZSM-5, and SAPO-34 membranes (Figure S7). A comparison of our results to other hybrid PSF membranes in literature is listed in Table 1.16,21−24,50,51 We find that, first, by using the designed dual-channel, molecular-sieving core/shell ZIF@MOF as fillers, the gas permeability and selectivity of asprepared hybrid PSF membrane can be positively improved simultaneously. Second, without sacrificing CO2/N2 selectivity our hybrid PSF membrane provides the largest increases in
g at 1 bar, Table S1), only ∼13% lower than the pristine MOF samples, suggesting that the internal large channels of MOF were still maintained for CO2. However, a significant reduction (∼73%) for N2 adsorption capacity was observed after five cycles of ZIF coating, dropping from 0.157 mmol/g (MOF) to 0.043 mmol/g (5-ZIF@MOF) at 1 bar. As a result, the ideal CO2/N2 adsorptive selectivity (the adsorbed capacity ratio of CO2 and N2 at 1 bar) dramatically increased from 17 to 54 (Table S1). We speculate that the improved CO 2 /N 2 adsorptive selectivity is attributed to the molecular-sieving effect of small pore of ZIF shell or smaller pores generated at the interface between ZIF and MOF surface. The TGA curves of MOF, ZIF-8, and 5-ZIF@MOF are shown in Figure S4. Under 100 °C, TGA showed significant mass loss from solvent evaporation. In the range of 100 to 250 °C, slight weight loss was observed for all the samples due to the desorbed water. With increased temperature, the MOF and ZIF@MOF samples will start to decompose at 350 °C, but pure ZIF will not decompose until 500 °C. By using the membrane preparation technique and controlling the layer-by-layer ZIF coating cycle numbers described previously, robust PSF membranes containing 40 wt % of MOF or ZIF@MOF were successfully fabricated. The morphology of as-prepared membranes was characterized by SEM. Figure 3a,d shows the cross-sectional SEM images of
Figure 3. Cross-section SEM images of PSF, PSF-MOF, and PSF-5ZIF@MOF membranes at lower magnification (a−c); cross-section SEM images of PSF, PSF-MOF, and PSF-5-ZIF@MOF membranes at higher magnification (d−f); water contact angles on PSF, PSF-MOF, and PSF-5-ZIF@MOF membranes (g−i).
pure PSF membrane at lower and higher magnifications, respectively. It can be seen that the pure PSF membrane is smooth and continuous. Figure 3b,c shows the thickness of asprepared PSF-MOF and PSF-5-ZIF@MOF membranes is highly uniform in the range of 60−70 μm with no sign of pinhole defects. In addition, strong interfacial interaction between fillers and polymer is observed in the membranes (Figure 3e,f). To further confirm favorable polymer/porous crystal interactions exist, differential scanning calorimetry measurements were carried out to study the Tg of membranes. Pure PSF membrane shows a Tg of 186.7 °C. The incorporation of MOF or ZIF@MOF has triggered a positive E
DOI: 10.1021/acs.nanolett.7b02910 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 4. Effect of layer-by-layer ZIF coating cycles on the pure gas permeability (a), and ideal gas selectivity (b) of hybrid PSF membranes. Pure gas permeation tests were measured at 3 bar and 35 °C. Error bars show standard deviations derived from triplicate measurements.
Table 1. Selected CO2 Permeability and CO2/N2 Selectivity % Increase of Hybrid PSF Membranes Reported in Literature and This Work
a
material
inorganic loading (wt %)
CO2 permeability increase (%)
CO2/N2 selectivity increase (%)
ref
PSF-ZIF@MOF PSF-UiO-66-NH2a PSF-13Xb PSF-MCM-48c PSF-SWNTd PSF-Silica PSF-ZIF-301a PSF-MIL-101a
40 50 33 20 15 20 30 24
710 770 −6 308 83 213 180 500
50 −13 310 −5 17 −33 12 44
this work Su et al.16 Gur21 Kim et al.22 Kim et al.23 Ahn et al.24 Sarfraz et al.50 Jeazet et al.51
Metal organic frameworks. bZeolite. cMesoporous silica. dSingle-walled nanotube.
■
CO2 permeability. We expect that the core/shell structural fillers can also be applied in principle to other challenging separation problems where no individual MOF structure is effective. Further, we anticipate that these core/shell architectures can be incorporated with other higher performing base polymers to fabricate more competitive hybrid membranes with enhanced separation performance. In conclusion, we demonstrate for the first time the formation of engineered core/shell structures of MOF, ZIF@ MOF, using large pore MOFs as the cores and small pore ZIFs as the shells, then apply the full core/shell particles as porous fillers in hybrid membranes. In this designed ZIF@MOF structure, the large pore MOF acts as a molecular transport highway and the small pore on the external surface works as a molecular filter. Compared with neat PSF polymer membranes, the optimized PSF-ZIF@MOF membranes showed remarkably improved CO2 separation performance (CO2 permeability of 45.2 barrer, CO2/N2 selectivity of 39) with 710% higher CO2 permeability and 50% higher CO2/N2 selectivity. Further, the as-prepared membrane exhibited improved surface hydrophobic property. We expect that the incorporation of advanced classes of porous crystals provides a new strategy to make novel hybrid membranes a promising avenue for CO2 capture.
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.7b02910. Synthesis of UiO-66-NH2 and ZIF-8, density measurements of hybrid membranes, and characterization of hybrid membranes (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Jinghua Guo: 0000-0002-8576-2172 Baoxia Mi: 0000-0003-3185-1820 Jeffrey J. Urban: 0000-0002-6520-830X Author Contributions ∥
Z.S. and F.Q. contributed equally to this work.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS Work at the Molecular Foundry, XRD data acquired by beamline 12.2.2 and WAXS data acquired at beamline 7.3.3. at the Advanced Light Source, Berkeley were supported by the Office of Science, Office of Basic Energy Sciences, of the U.S. F
DOI: 10.1021/acs.nanolett.7b02910 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
(32) Ordoñez, M. J. C.; Balkus, K. J., Jr; Ferraris, J. P.; Musselman, I. H. J. Membr. Sci. 2010, 361, 28−37. (33) Dai, Y.; Johnson, J. R.; Karvan, O.; Sholl, D. S.; Koros, W. J. J. Membr. Sci. 2012, 401−402, 76−82. (34) Zhang, C.; Lively, R. P.; Zhang, K.; Johnson, J. R.; Karvan, O.; Koros, W. J. J. Phys. Chem. Lett. 2012, 3, 2130−2134. (35) Song, Z.; Huang, Y.; Xu, W. W.; Wang, L.; Bao, Y.; Li, S. G.; Yu, M. Sci. Rep. 2015, 5, 13981. (36) Song, Z.; Huang, Y.; Wang, L.; Li, S.; Yu, M. Chem. Commun. 2015, 51, 373−375. (37) Yu, M.; Funke, H. H.; Noble, R. D.; Falconer, J. L. J. Am. Chem. Soc. 2011, 133, 1748−1750. (38) Kim, J. H.; Ishida, A.; Okajima, M.; Niwa, M. J. Catal. 1996, 161, 387−392. (39) Huang, A.; Bux, H.; Steinbach, F.; Caro, J. Angew. Chem. 2010, 122, 5078−5081. (40) Huang, K.; Dong, Z.; Li, Q.; Jin, W. Chem. Commun. 2013, 49, 10326−10328. (41) Shekhah, O.; Eddaoudi, M. Chem. Commun. 2013, 49, 10079− 10081. (42) Tang, J.; Dong, W.; Wang, G.; Yao, Y.; Cai, L.; Liu, Y.; Zhao, X.; Xu, J.; Tan, L. RSC Adv. 2014, 4, 42977−42982. (43) Shen, L.; Liang, S.; Wu, W.; Liang, R.; Wu, L. Dalton Trans. 2013, 42, 13649−13657. (44) Yang, J.; Dai, Y.; Zhu, X.; Wang, Z.; Li, Y.; Zhuang, Q.; Shi, J.; Gu, J. J. Mater. Chem. A 2015, 3, 7445−7452. (45) Tuninetti, J. S.; Rafti, M.; Azzaroni, O. RSC Adv. 2015, 5, 73958−73962. (46) Huang, Y.; Qin, W.; Li, Z.; Li, Y. Dalton Trans. 2012, 41, 9283− 9285. (47) Cacho-Bailo, F.; Matito-Martos, I.; Perez-Carbajo, J.; EtxeberriaBenavides, M.; Karvan, O.; Sebastian, V.; Calero, S.; Tellez, C.; Coronas, J. Chem. Sci. 2017, 8, 325−333. (48) Shen, J.; Liu, G.; Huang, K.; Li, Q.; Guan, K.; Li, Y.; Jin, W. J. Membr. Sci. 2016, 513, 155−165. (49) Song, Q.; Nataraj, S. K.; Roussenova, M. V.; Tan, J. C.; Hughes, D. J.; Li, W.; Bourgoin, P.; Alam, M. A.; Cheetham, A. K.; AlMuhtaseb, S. A.; Sivaniah, E. Energy Environ. Sci. 2012, 5, 8359−8369. (50) Sarfraz, M.; Ba-Shammakh, M. J. Membr. Sci. 2016, 514, 35−43. (51) Jeazet, H. B.; Koschine, T.; Staudt, C.; Raetzke, K.; Janiak, C. Membranes 2013, 3, 331−353.
Department of Energy under Contract Number DE-AC0205CH11231.
■
REFERENCES
(1) Change, I. P. O. C. Agenda 2007, 6, 333. (2) Haszeldine, R. S. Science 2009, 325, 1647−1652. (3) Orr, J. C.; Fabry, V. J.; Aumont, O.; Bopp, L.; Doney, S. C.; Feely, R. A.; Gnanadesikan, A.; Gruber, N.; Ishida, A.; Joos, F.; Key, R. M.; Lindsay, K.; Maier-Reimer, E.; Matear, R.; Monfray, P.; Mouchet, A.; Najjar, R. G.; Plattner, G.-K.; Rodgers, K. B.; Sabine, C. L.; Sarmiento, J. L.; Schlitzer, R.; Slater, R. D.; Totterdell, I. J.; Weirig, M.-F.; Yamanaka, Y.; Yool, A. Nature 2005, 437, 681−686. (4) Wang, J.; Huang, L.; Yang, R.; Zhang, Z.; Wu, J.; Gao, Y.; Wang, Q.; O’Hare, D.; Zhong, Z. Energy Environ. Sci. 2014, 7, 3478−3518. (5) McDonald, T. M.; Lee, W. R.; Mason, J. A.; Wiers, B. M.; Hong, C. S.; Long, J. R. J. Am. Chem. Soc. 2012, 134, 7056−7065. (6) Kintisch, E. Science 2007, 317, 184−186. (7) Liu, L.; Qiu, W.; Sanders, E. S.; Ma, C.; Koros, W. J. J. Membr. Sci. 2016, 510, 447−454. (8) Mason, J. A.; McDonald, T. M.; Bae, T.-H.; Bachman, J. E.; Sumida, K.; Dutton, J. J.; Kaye, S. S.; Long, J. R. J. Am. Chem. Soc. 2015, 137, 4787−4803. (9) McDonald, T. M.; Lee, W. R.; Mason, J. A.; Wiers, B. M.; Hong, C. S.; Long, J. R. J. Am. Chem. Soc. 2012, 134, 7056−7065. (10) Zhang, Z.; Yao, Z. Z.; Xiang, S.; Chen, B. Energy Environ. Sci. 2014, 7, 2868−2899. (11) Lin, L.-C.; Berger, A. H.; Martin, R. L.; Kim, J.; Swisher, J. A.; Jariwala, K.; Rycroft, C. H.; Bhown, A. S.; Deem, M. W.; Haranczyk, M.; et al. Nat. Mater. 2012, 11, 633−641. (12) Rochelle, G. T. Science 2009, 325, 1652−1654. (13) Rinker, E. B.; Ashour, S. S.; Sandall, O. C. Ind. Eng. Chem. Res. 2000, 39, 4346−4356. (14) Darunte, L. A.; Walton, K. S.; Sholl, D. S.; Jones, C. W. Curr. Opin. Chem. Eng. 2016, 12, 82−90. (15) Fan, Y.; Kalyanaraman, J.; Labreche, Y.; Rezaei, F.; Lively, R. P.; Realff, M. J.; Koros, W. J.; Jones, C. W.; Kawajiri, Y. Ind. Eng. Chem. Res. 2015, 54, 1783−1795. (16) Su, N. C.; Sun, D. T.; Beavers, C. M.; Britt, D. K.; Queen, W. L.; Urban, J. J. Energy Environ. Sci. 2016, 9, 922−931. (17) Adatoz, E.; Avci, A. K.; Keskin, S. Sep. Purif. Technol. 2015, 152, 207−237. (18) Kang, Z.; Peng, Y.; Qian, Y.; Yuan, D.; Addicoat, M. A.; Heine, T.; Hu, Z.; Tee, L.; Guo, Z.; Zhao, D. Chem. Mater. 2016, 28, 1277− 1285. (19) Morigami, Y.; Kondo, M.; Abe, J.; Kita, H.; Okamoto, K. Sep. Purif. Technol. 2001, 25, 251−260. (20) Seoane, B.; Coronas, J.; Gascon, I.; Benavides, M. E.; Karvan, O.; Caro, J.; Kapteijn, F.; Gascon, J. Chem. Soc. Rev. 2015, 44, 2421− 2454. (21) Gür, T. M. J. Membr. Sci. 1994, 93, 283−289. (22) Kim, S.; Marand, E.; Ida, J.; Guliants, V. V. Chem. Mater. 2006, 18, 1149−1155. (23) Kim, S.; Chen, L.; Johnson, J. K.; Marand, E. J. Membr. Sci. 2007, 294, 147−158. (24) Ahn, J.; Chung, W. J.; Pinnau, I.; Guiver, M. D. J. Membr. Sci. 2008, 314, 123−133. (25) Bétard, A.; Fischer, R. A. Chem. Rev. 2012, 112, 1055−1083. (26) Zhang, C.; Koros, W. J. J. Phys. Chem. Lett. 2015, 6, 3841−3849. (27) Yeo, Z. Y.; Chai, S.-P.; Zhu, P. W.; Mohamed, A. R. RSC Adv. 2014, 4, 54322−54334. (28) Ahmadi Feijani, E.; Tavasoli, A.; Mahdavi, H. Ind. Eng. Chem. Res. 2015, 54, 12124−12134. (29) Denny, M. S., Jr; Moreton, J. C.; Benz, L.; Cohen, S. M. Nat. Rev. Mater. 2016, 1, 16078. (30) Li, H.; Song, Z.; Zhang, X. J.; Huang, Y.; Li, S. G.; Mao, Y. T.; Ploehn, H. J.; Bao, Y.; Yu, M. Science 2013, 342, 95−98. (31) Bae, T. H.; Lee, J. S.; Qiu, W.; Koros, W. J.; Jones, C. W.; Nair, S. Angew. Chem., Int. Ed. 2010, 49, 9863−9866. G
DOI: 10.1021/acs.nanolett.7b02910 Nano Lett. XXXX, XXX, XXX−XXX
