Zeyu Lia,
Zhejian Caob,
Carlos Grande
c,
Wenjing Zhang
d,
Yibo Dou
d,
Xiaosen Li
e,
Juan Fue,
Nasir Shezadb,
Farid Akhtar
b and
Andreas Kaiser
*a
aDepartment of Energy Conversion and Storage, Technical University of Denmark (DTU), Denmark. E-mail: akai@dtu.dk
bDepartment of Engineering Sciences and Mathematics, Luleå University of Technology (LTU), Sweden
cProcess Technology, SINTEF, Norway
dDepartment of Environmental Engineering, Technical University of Denmark (DTU), Denmark
eGuangzhou Center for Gas Hydrate Research, Chinese Academy of Sciences (CAS), China
First published on 23rd December 2021
Polyacrylonitrile (PAN) nanofibers were prepared by electrospinning and coated with zeolitic imidazolate framework-8 (ZIF-8) by a phase conversion growth method and investigated for CO2 capture. The PAN nanofibers were pre-treated with NaOH, and further coated with zinc hydroxide, which was subsequently converted into ZIF-8 by the addition of 2-methyl imidazolate. In the resulting flexible ZIF-8/PAN composite nanofibers, ZIF-8 loadings of up to 57 wt% were achieved. Scanning electron microscopy and energy-dispersive X-ray spectroscopy (EDS) showed the formation of evenly distributed submicron-sized ZIF-8 crystals on the surface of the PAN nanofibers with sizes between 20 and 75 nm. X-ray photoelectron spectroscopy (XPS) and carbon-13 nuclear magnetic resonance (13C NMR) investigations indicated electrostatic interactions and hydrogen bonds between the ZIF-8 structure and the PAN nanofiber. The ZIF-8/composite nanofibers showed a high BET surface area of 887 m2 g−1. CO2 adsorption isotherms of the ZIF-8/PAN composites revealed gravimetric CO2 uptake capacities of 130 mg g−1 (at 298 K and 40 bar) of the ZIF-8/PAN nanofiber and stable cyclic adsorption performance.
Beneath shaping of microporous materials into useful structures, such as pellets, granules or extrudates,30 recently, new approaches have been proposed to structure adsorbents (carbons, zeolites, MOFs etc.) into nanofibers to enhance gas transport,17 adsorption capacity18 and adsorption kinetics.19 For example, zeolite nanofibers of ZSM-5 have been fabricated by electrospinning, post shaping and carbonization of polyvinylpyrrolidone (PVP)20 the performance has been evaluated for CO2 separation from CO2/CH4 mixtures in synthetic biogas mixtures.
Due to the temperature instability of MOF materials, this approach to fabricate MOF-nanofiber composite materials is not feasible for the majority of MOF-based nanofibers. However, different approaches for the structuring of MOF-nanofiber materials for energy and environmental applications have been recently reviewed by Dou et al.,21 including in situ growth of MOF nanocrystals on the surface of polymer nanofibers. The preparation of MOF-polymer nanofibers through electrospinning is an elegant way to shape MOF materials into hierarchical porous nanofibrous sorbent materials at mild conditions with large surface-to-volume ratio, tailored pore size, and high permeability for gas separation processes. In this study we propose a phase inversion method to grow ZIF-8 on the surface of electrospun polyacrylonitrile (PAN) nanofibers for CO2 capture. We investigated the influence of the growth of the ZIF-8 nanoparticles on PAN nanofibers as a function of reaction time to maximize the CO2 gas uptake, which was expected to increase with the amount of the microporous ZIF-8 sorbent loaded on the inactive PAN nanofiber as backbone.
The resulting ZIF-8/PAN composite nanofibers have been characterized by scanning electron microscope-energy dispersive X-ray spectroscopy (SEM-EDS), X-ray diffraction pattern (XRD), surface area and pore size distribution analysis and by more advanced characterization techniques, such as X-ray photoelectron spectroscopy (XPS) and nuclear magnetic resonance spectroscopy (NMR).
Finally, the CO2 uptake of the ZIF-8/PAN composite nanofibers have been tested at pressures up to 40 bars and the cyclic adsorption–desorption properties of the composites have been investigated.
For the preparation of the PAN nanofibers, 1 g PAN and 10 g DMF were mixed and sealed into a 50 mL plastic bottle with cylindrical zirconia balls to ball-mill by regular ball milling at 50 rpm for 24 hours. Then the dispersion solution was electrospun by needle-based electrospinning equipment (Linari s.r.l) with the voltage of 30 kV and a distance of 140 mm between the electrospinning needle and receiving substrate to prepare PAN nanofibers.
For the activation of the PAN nanofiber surface with hydroxide groups, 0.5 g of PAN nanofibers were immerged into the 50 mL NaOH-D.I.W solution (60 g NaOH dissolved into 500 mL D.I.W) and heated at 75 °C for 20 min to activate the fiber surface. The treated PAN nanofibers were 3 times washed with ethanol and dried afterwards in air at room temperature.
For the application of the metal precursor, the surface-treated PAN nanofibers were immersed into 100 mL of a 120 millimole per liter (mM) Zn(NO3)2 solution for 20 min to form Zn(OH)2 on the surface of PAN nanofibers. The applied Zn(NO3)2 solution consisted of 6.29 g Zn(NO3)2·6H2O in 180 mL of a 1:
1 volumetric solvent mixture of methanol and ethanol.
For the final in situ growth process, 100 mL 480 mM 2-mim solution was added to the Zn(NO3)2 solution directly, where the Zn(OH)2-coated PAN fiber had been immerged already to complete the in situ growth process. Every 30 minutes, the samples were taken and dried after ethanol washing to identify the optimum loading and structure of ZIF-8 in the composite. The applied 2-mim solution consisted of 7.49 g 2-methylimidazole in 190 mL of a 1:
1 volumetric solvent mixture of methanol and ethanol. According to reaction time the samples were labelled ZIF-8/PAN-X (with X resembling 30, 60 and 90 minutes of reaction time, see Table 1 further below).
Sample ID | Average crystal size (nm) | Average pore size (Å) | BET surface area (m2 g−1) | Total pore volume (cm3 g−1) | Micro pore volumea (cm3 g−1) |
---|---|---|---|---|---|
a The micro-pore volumes have been calculated from a T-plot, following a method reported by Galarneau et al.22 | |||||
ZIF-8 powder | 101 | 13.8 | 1015 | 0.70 | 0.51 |
ZIF-8/PAN-30 | 20 | 15.2 | 495 | 0.38 | — |
ZIF-8/PAN-60 | 36 | 13.0 | 862 | 0.56 | — |
ZIF-8/PAN-90 | 76 | 11.2 | 888 | 0.50 | 0.31 |
X-ray diffraction (XRD) measurements were performed by a Rigaku Smartlab X-ray diffractometer, which was operated at 40 kV and 30 mA using Cu Kα radiation with step speed of 5° min−1. All of the samples were attached on a silicon-substrate sample holder. The miller indexes were identified by Match! 3 software.
X-ray photoelectron spectroscopy (XPS) was performed in the range of 200 to 1200 eV with ZIF-8 nanocrystals, pure PAN fiber and ZIF-8/PAN composite nanofibers in a angle resolved XPS chamber for depth profiling of surfaces (equipment from Thermo Scientific Nexsa).
13C nuclear magnetic resonance spectroscopy (NMR) was performed on a Bruker Avence III 300 Wide Bore solid state NMR spectrometer, using a 4 mm probe. The frequency of 13C was 75 MHz with the pulse program of CP-Toss, using 4 mm rotors at a MAS spinning frequency of 6500 Hz. The 90° 13C pulse-length was 4 μs, and pulse length in decoupling sequence is 7. Recycle delay is 5 s, while the sweep wide is 30241.94 Hz. The carbon resonance line of adamantane was used as an external chemical shift standard and was assigned a value of 38.48 ppm. The scans for CP-Toss experiments was 1k. Besides, the samples were frozen and milled into powder with help of liquid nitrogen before it was filled into NMR tube.
The surface area as well as pore-related information (size distribution, total pore volume, micro-pore volume) determination of ZIF-8/PAN composite materials and ZIF-8 powders were measured by nitrogen adsorption isotherms at −196 °C (Autosorb-iQ-2, specific analysis for microstructure). The CO2 uptake measurements were also carried out in this equipment. All samples were degassed to below 10−6 bar and 383 K for 8 hours prior to the measurement.
The high-pressure CO2 adsorption performance was measured by an IsoSORP adsorption analyzer (TA instruments, New Castle, DE, USA), on samples with a weight of about 0.5 g. All samples were degassed at 110 °C under high vacuum for 4 h, following a buoyancy test with helium at 22 °C to determine the mass and volume of each sample. Then, the CO2 adsorption–desorption isotherms of the obtained ZIF-8/PAN composite were measured with CO2 dosing from high vacuum to 40 bar, then to high vacuum with an equilibrium point every 10 bar. The equilibrium for each step was set until the deviation of mass was less than 0.1 mg per 10 min.
The pattern-ray diffractogram of the synthesized ZIF-8/PAN composite nanofibers matched closely with the simulated X-ray diffractogram of ZIF-8 powder, which can be assigned to cubic I3m space group,15 confirming the growth of ZIF-8 nanoparticles on the surface of PAN nanofibers. Lower relative diffraction intensity of X-rays and broadening of diffraction peaks at FWHM (full width at half maximum) can be observed for the ZIF-8/PAN composites samples with decreasing reaction time, Fig. 2. This suggests that the crystallinity and crystallite size of the ZIF-8 particles, attached to the PAN nanofibers, increases with reaction time. Crystallite sizes increase from 20 to 70 nm for increasing reaction time between 30 to 90 minutes. Effects of crystallite growth and size in ZIF-8 nanopowder has previously been investigated by Tanaka et al.26
Table 1 summarizes the average crystal size of the ZIF-8 crystals along with the surface textural properties of the ZIF-8/PAN composite nanofiber and the ZIF-8 nanopowder, which are discussed in Section 3.3.
Fig. 3a and c show the SEM micrographs and Fig. 3b and d the Zinc elemental mappings of the Zn(OH)2/PAN composite nanofibers (zinc hydroxide attached to the PAN nanofiber before the in situ growth of the MOF) and of the ZIF-8/PAN composite-90 nanofiber structure. The larger size and broader distribution of the Zinc in the initial Zn(OH)2/PAN composite (Fig. 3b) compared to the final ZIF-8/PAN-90 composite nanofiber (Fig. 3d) indicate that there occurs a redistribution of zinc atoms from the initial Zn(OH)2 phase during the phase conversion process into the ZIF-8 crystallites. The explanation is that during the growth of the ZIF-8 on the surface of nanofiber, zinc ions from the hydroxide phase are finally incorporated on a molecular level into the crystal structure of the MOF particles, which results in a more even and finer coverage of the PAN nanofiber surface with Zn.
To verify interaction between the PAN nanofiber and ZIF-8 framework, narrow scans of the N 1s and C 1s were analyzed27 for the ZIF-8/PAN-90 composite nanofiber, the ZIF-8 powder and pure PAN nanofibers, respectively (Fig. 4). The energy peak of the N 1s exhibits a 0.80 eV red shift from 400.8 to 399.3 eV after in situ growth of ZIF-8 on the PAN nanofiber. Similarly, the C 1s signal shows a red shift of 1.9 eV after the PAN nanofiber has been coated with ZIF-8 and resulting in ZIF-8/PAN-90 composite. These shifts are related to the change of the atomic chemical environment of carbon and nitrogen, indicating interaction between the PAN nanofiber and the ZIF-8 structure.
13C NMR spectroscopy was conducted to further analyse the atomic chemical environment. As shown in Fig. 5, the location of the resonance peaks at 152.0 and 14.6 ppm, respectively, can be assigned to protonated aromatic carbon atoms (C1, carbon) and the methyl group (C3, –CH3). These peaks have no major chemical shift.
However, there is a visible chemical shift by 1.7 ppm from 123.7 ppm to 125.4 ppm for the double-bond carbon (C2, –CHCH–) when comparing the spectra for the ZIF-8 nanopowder with the ZIF-8 crystallites attached to the ZIF-8/PAN-90 composite nanofiber. Similar chemical shift of the C2 peak in ZIF-8 has been observed when organic molecules, such as caffeine, were encapsulated in the pores of ZIF-8 to form van der Waals bonds.28 However, more detailed analysis supported by DFT calculations29 would be required to further proof electrostatic interactions and hydrogen bonding between the ZIF-8 and C–N groups of the PAN nanofiber (as visualized by the chemical formulas in Fig. 5).
A large peak centered around a pore radius of about 11 Å can be assigned to micropores, whereas two smaller and broader peaks between 20 to 70 A are related to micro and mesopores. Table 1 summarizes the micropore and total pore volume and BET surface area of all ZIF-8/PAN nanofibers and the ZIF-8 nanopowder. The comparably low total pore volume of 0.38 cm3 g−1 and specific surface area (495 m2 g−1) of the sample obtained after short reaction time (30 minutes), indicate that the ZIF-8 has not been fully crystallized or has major surface defects. The broad and slightly displaced XRD peaks for this sample in Fig. 2 support this thesis. After longer reaction time (90 minutes), the total pore volume increases to 0.51 cm3 g−1 and the surface area to 888 m2 g−1 for the sample ZIF-8/PAN-90. Considering a maximum ZIF-8 loading of 57.19 wt% in ZIF-8/PAN-90 achieved in this work, the surface area of the pure ZIF-8 crystallites in the composite can be calculated to 1552 m2 g−1, assuming negligible surface area of the PAN nanofibers. Using similar estimates (43 wt% PAN in the composite), the actual micropore and total pore volumes of the ZIF-8 crystallites in the ZIF-8/PAN-90 composite should be about 0.54 and 0.88 cm3 g−1, respectively. This indicates that by the in situ growth method proposed in this work, ZIF-8/PAN nanofiber composites with high surface area, pore volume and ZIF-8 particle loadings have been achieved compared to ZIF-8-nanofibers in other studies. Table 2, further below, gives a comparision of the CO2 uptake performance of the ZIF-8 nanofiber composite materials in this study compared to other studies.
Material | ZIF-8 loading (wt%) | BET surface area (m2 g−1) | CO2 uptakea (cm3 g−1) | Reference |
---|---|---|---|---|
a CO2 uptake values at low pressures are given at 25 °C, 1 bar and in cm3 g−1. If CO2 uptake was measured under other conditions, for example higher pressures or lower temperature, these values are given in brackets.b The ZIF-8 loading unknown. | ||||
ZIF-8/PAN-90 | 57.2 | 888 | 7.0 | This work |
ZIF-8 powder | 100 | 1016 | 14.7 | This work |
ZIF-8/PAN-90 | 57.2 | 888 | 130 (mg g−1, 40 bar) | This work |
ZIF-8 powder | 100 | 1813 | 374 (mg g−1, 40 bar) | Autié et al.11 |
ZIF-8/ZnO core–shell | — | 733 | 7.6 | Thomas et al.31 |
ZIF-8/PAN | —b | 983 | 13.3 (20 °C) | Gao et al.32 |
ZIF-8 powder | 100 | 880 | 16.5 | Gao et al.33 |
ZIF-8 powder | 100 | — | 15.3 | Huang et al.34 |
ZIF-8 powder | 100 | 1264 | 350 (mg g−1, RT, 30 bar) | Nune et al.35 |
Fig. 7 shows the cyclic CO2 uptake and re-generation of the ZIF-8/PAN-90 at 25 °C for 1 bar and for 40 bars for four cycles in Fig. 7a and b, respectively. The CO2 uptake of ZIF-8/PAN-90 nanofibers at 1 bar and 40 bars is 7 cm3 g−1 and 130 mg g−1.
![]() | ||
Fig. 7 (a) Cyclic adsorption isotherms up to 1 bar at 25 °C and (b) the cyclic CO2 uptake of ZIF-8/PAN-90 composite at high pressure of 40 bars. |
These initial results demonstrate good initial cyclic stability for CO2 separation with fast adsorption–desorption kinetics. Reasonable high CO2 uptake of ZIF-8/PAN-90 nanofiber composites have been achieved compared to similar studies on ZIF-8,36 suggesting that these materials are interesting for CO2 separation from gas streams.
This journal is © The Royal Society of Chemistry 2022 |