Highly CO2 perm-selective metal-organic framework membranes through CO2 annealing post-treatment

Journal of Membrane Science 555 (2018) 97–104

Contents lists available at ScienceDirect

Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci

Highly CO2 perm-selective metal-organic framework membranes through CO2 annealing post-treatment Zebao Ruia, Joshua B. Jamesb, Y.S. Linb, a b

T

⁎

School of Chemical Engineering and Technology, Sun Yat-sen University, Zhuhai 519082, PR China Chemical Engineering, School for Engineering of Matter, Transport and Energy, Arizona State University, Tempe, AZ 85287, USA

A R T I C LE I N FO

A B S T R A C T

Keywords: Metal-organic framework membrane Gas separation Adsorption-driven selectivity Post-treatment Carbon dioxide

Gas separation by metal-organic framework (MOF) membranes are still unsatisfactory due partly to unsatisfied separation characteristics caused by the trade-off between selectivity and permeability. Herein, we provide a facile post CO2 annealing method to remarkably improve both the permeance and separation factor of IRMOF-1 membranes for CO2 separation from CO2/H2 mixture. Post-treatment of the IRMOF-1 membrane by annealing at 100 °C under a high pressure CO2 stream deceases H2 permeance while increasing CO2 permeance, leading to simultaneous enhancement in both CO2/H2 separation factor from 721 up to 5781 and CO2 permeance from 5.67 × 10−7 up to 9.38 × 10−7 mol m−2 s−1 Pa−1 at CO2 molar fraction of 98%, feed pressure of 5 atm and 298 K. The unusual separation behavior is related to the enhanced CO2 adsorption selectivity over H2 and reduced CO2 affinity of the membrane due to the formation of surface carbonate anions caused by the CO2 treatment. The proposed gas atmosphere post-treatment strategy opens a window for designing next generation MOF-related gas separation membranes.

1. Introduction Separation of carbon dioxide from H2, CH4 and N2 containing gases is important with respect to environmental sustainability and energy utilization [1–4]. Among various separation processes, the CO2 selective membrane is particularly attractive in concentrating valuable commodities such as H2 and CH4 at high pressures [2–4] and producing high grade CO2 [3]. Among these gas mixtures, the greatest challenge is the separation of CO2 and H2 due to the similar kinetic diameters of H2 (~ 0.29 nm) and CO2 (~ 0.33 nm) [4]. To date, CO2-selective membranes are dominated by rubbery polymeric membranes owing to their easy processability and scalability [4–6]. Nonetheless, polymeric membranes normally suffer from plasticization [7], which is of particular concern for CO2 enriched feed stream, and the trade-off between permeability and selectivity [5,6]. Alternatively, microporous membranes, especially zeolite membranes, have been extensively studied for CO2 separation from CO2/H2 [4,8–11]. However, the majority of zeolite membranes only show a good CO2/H2 separation performance at low temperatures and deteriorates at a high feed pressure [4]. These membranes also suffer from the selectivity/permeability trade-off, restricting their practical application [4]. In consideration with the facilely tunable structure properties and tailorable surface functionalities of metal-organic framework (MOF)

⁎

materials [12–17], MOF membranes provide the choice to simultaneously achieve good selectivity and high permeability. For CO2-selective MOF membranes, Takamizawa et al. [18] found that singlecrystal [Cu2(benzoate)4(pyrazine)] MOF membrane was CO2 perm-selective over H2. Zhao and Lin [19] observed a slight increase in CO2/H2 selectivity with increasing feed pressure and obtained a CO2/H2 selectivity close to 5 at ~ 3.4 bar, 298 K and a CO2 feed composition of ~ 82% in MOF-5 membrane. Al-Maythalomy et al. [2] synthesized a continuous sod-ZMOF membrane with CO2 adsorption-driven selectivity, which displayed a CO2/H2 selectivity of 5.2 at 308 K and 3.4 bar. Despite the great potential and fast development of MOF membranes, state of the art CO2-selective MOF membranes generally show poorer performance than those high-quality zeolite and polymer membranes. Post-synthetic modification is an effective way to tune the physical and chemical properties of MOFs [20] and opens a window to enhance their separation performance [14,15,21]. However, both the complex modification process and undesirable trade-off between permeability and selectivity by post-functionalization with chemical reagents restrict its application [21]. It is still a great challenge to break such an unwanted permeability-selectivity trade-off and develop a powerful MOF membrane for CO2-selective separation. Our strategy is to enable enhancement in CO2 permeation flux by

Corresponding author. E-mail address: [email protected] (Y.S. Lin).

https://doi.org/10.1016/j.memsci.2018.03.036 Received 15 January 2018; Received in revised form 15 March 2018; Accepted 16 March 2018 Available online 16 March 2018 0376-7388/ © 2018 Elsevier B.V. All rights reserved.

Journal of Membrane Science 555 (2018) 97–104

Z. Rui et al.

2.2. Characterization

employing a MOF with a relatively large aperture size and simultaneous promotion in CO2 separation factor by tuning MOF pore wall chemistry. IRMOF-1 (or MOF-5), is the first robust MOF [12] and exhibits an extremely large empty cavity (good for CO2 permeability), high surface area and exceptionally high CO2 uptake (good for CO2 adsorptiondriven selectivity) [12,22,23], making it a preferred model MOF material for constructing a membrane for CO2 separation with CO2 adsorption-driven selectivity [3,24]. Herein, high quality IRMOF-1 (or MOF-5) membranes supported on an α-Al2O3 support with a thickness of ~ 14 µm was successfully synthesized by a secondary growth method [3], and then post-modified under high pressure CO2 atmosphere (~5 atm) and relatively high temperature (~ 100 °C, referred as PMOF5) for the CO2 separation from CO2-enriched CO2/H2 gas mixtures. The PMOF-5 membrane exhibits a CO2/H2 separation factor as high as ~ 5781 and a high CO2 permeance up to ~ 9.38 × 10−7 mol m−2 s−1 Pa−1 (about 2800 GPU) at the optimized feed conditions and 298 K. The unusual separation behavior of PMOF-5 membrane is related to the enhanced CO2 uptake over H2 and weakened CO2 affinity of the membrane caused by the CO2 treatment. The demonstrated exceptional separation properties of PMOF-5 membranes break the selectivity/ permeability trade-off limitation for MOF membrane-related gas separation and offers a new method for the development of adsorption controlled CO2 separation membranes.

The surface morphology of the MOF-5 membranes was examined by a Philips FEI XL-30 scanning electron microscope at accelerating voltage of 20 kV. The crystal phase structure of the samples was examined by X-ray diffraction (XRD) using a conventional Bruker D8 diffractometer using CuKα (λ = 0.1543 nm) radiation with a scan speed of 2°/min. Fourier transform infrared spectra (FTIR) were obtained with a Thermo Scientific Nicolet iS50 FTIR Spectrometer in the frequency range of 4000–400 cm−1. In situ Diffuse Reflectance Infrared Fourier Transformed Spectroscopy (DRIFTS) analysis was conducted on a Thermo Scientific Nicolet iS50 Fourier Transform Infrared (FTIR) equipped with a DTGS potassium bromide (KBr) detector and diffuse reflectance. About 10 mg finely ground MOF-5 or PMOF-5 powders were packed in the in situ chamber and mixed with approximately 50 mg of KBr to prevent detector saturation. 100 mL/min of CO2 was introduced. The spectra under adsorption conditions were recorded after 64 scans with a resolution of 4 cm−1. The in situ DRIFT spectrum under vacuum with pure KBr at room temperature was measured and taken as a background for each sample. The pure component adsorption isotherms of CO2 at three temperatures (273, 298, and 323 K) and H2 at 273 K at various gas pressures up to 1 atm were measured volumetrically for MOF-5 and PMOF-5 powders using a Micromeritics ASAP 2020 adsorption apparatus. The temperatures were controlled by a dewar with a circulating jacket connected to a thermostatic bath with a precision of ± 0.01 °C. The degas procedure was repeated in all samples before measurements at 100 °C for 12 h. N2 adsorption was performed at 77 K to determine the surface area and pore size of the samples.

2. Experimental 2.1. Powders and membrane preparation MOF-5 powders were synthesized via the solvothermal method reported in literature with some modifications [19,24]. 2.80 mmol zinc nitrate hexahydrate (Zn(NO)3·6H2O, 99%, Sigma Aldrich) and 1.06 mmol terephthalic acid (BDC, 99+%, Acros Organics) were dissolved in 40 mL 2,2-dimethylformamide (DMF, 99.8+%, Alfa Aesar) in a 100 mL glass vial. The solution was then stirred vigorously to allow for the precursors to completely dissolve while a given amount of Nethyldiisopropylamine (EDIA, 0.5 mL, +99.5%, Acros Organic) was slowly added drop-wisely. The glass vial was then sealed, heated to 403 K and held for 4 h under autogenous pressure by solvothermal synthesis. After the reaction, the vial was taken out of the oil bath and cooled down to the room temperature naturally. The cubic-like crystals of colorless powder were collected from the solvent by filtration and washed subsequently with DMF and chloroform to remove the unreacted zinc nitrate. After that, the crystals were immersed in chloroform (50 mL), sealed tightly and put into the oven at 343 K for three days. During the heating process, the solvent was decanted and replenished each day. Finally, the sample was dried under vacuum at 373 K overnight. The as-prepared MOF-5 crystals were used for CO2 adsorption, post-treatment and membrane preparation. Fine MOF-5 seed particles were dispersed in DMF solution at a concentration of ~ 1.5 wt% and ultrasonically agitated for 4 h to create a well-dispersed MOF-5 suspension. Once a stable MOF-5 suspension was obtained, the membrane synthesis via dip-coating and secondary growth proceeded over a porous α-Al2O3 disks with thickness of 2 mm and diameter of 20 mm [3]. Before characterization and gas permeation tests, the as-synthesized MOF-5 membranes were dried at 373 K overnight under vacuum. Post CO2 atmosphere treatment of MOF-5 particles (PMOF-5) were performed at 100 °C and 5 atm for 15 h. For the MOF-5 membrane, the post treatment was conducted in a steady state multicomponent gas permeation/separation system, as depicted in the previous work [19]. A MOF-5 membrane sample was mounted in a stainless-steel membrane cell, with the MOF-5 membrane layer on the feed side, and sealed by silicone O-rings. The permeation area of the membrane was 2.24 × 10−4 m2. The membrane was post-treated at 100 °C for 15 h with 5 atm pure CO2 at the feed side and the permeate side connecting with the bubble flowmeter.

2.3. Gas permeation/separation tests The gas permeation tests for single gas CO2 or H2, or a mixture of CO2/H2 at various compositions, a feed pressure of 5 atm and various temperatures were performed using the permeation/separation system. The composition of the gas mixture in the feed was controlled by two mass flow controllers. The pressure of the feed was controlled by using a needle valve on the retentate line. The gas compositions at the feed or permeate side were analyzed by gas chromatography (GC) (Alltech Haysep DB 100/120 column, TCD detector, and argon as the carrier gas). The permeance and separation factor are defined and calculated as

Fi =

Qp yi S (Pf x i − Pp y) i

α (i/ j ) =

(1)

yi / yj x i /xj

(2)

where Qp is the molar flow rate of the permeate side (measured by bubble flow meter), yi and xi are molar fraction for species i in permeate and retentate streams measured by GC, Pf and Pp are the total pressure in the feed or permeate side measured by a pressure meter, and S is the membrane permeation area. 3. Results and discussion 3.1. Effect of CO2 post-treatment on MOF membranes MOF-5 membranes with a thickness about 14 µm were fabricated over an α-Al2O3 support by the secondary growth method, as demonstrated by the surface morphology (Fig. 1a) and systematical characterizations in our previous work [3]. After post-treatment under 5 atm on-stream CO2 atmosphere at 100 °C (PMOF-5) for 15 h, no obvious differences in surface morphology between the as-prepared MOF5 and PMOF-5 membranes are observed (Fig. 1a&b). MOF-5 crystals of about 5–20 µm in size are dispersed randomly over both MOF-5 and 98

Journal of Membrane Science 555 (2018) 97–104

Z. Rui et al.

(220)

Fig. 1. SEM images of the (a) as-prepared MOF-5 and (b) post-treated MOF-5 (or PMOF-5) membranes.

(200) (400) (420)

(220) (200)

(b) (400) (420)

(a)

PMOF-5

PMOF-5

MOF-5

MOF-5

a-Al 2O3 10

20

30

40

10

50

20

30

40

50

2-Theta

2-Theta

Fig. 2. XRD patterns: (a) MOF-5 membrane, PMOF-5 membrane and Al2O3 support; (b) MOF-5 and PMOF-5 powders.

temperatures and a feed pressure of 5 atm. As shown in Fig. 3, both the CO2 and H2 permeances decrease with increasing CO2 mole fraction ( x CO2 ) in the feed, and the rate of decrease in H2 permeance is larger than that of CO2, resulting in an increase in the separation factor α(CO2/H2) according to the increase in x CO2 for both membranes. The α(CO2/H2) of the MOF-5 membrane increases drastically from ~2 at a x CO2 below 0.94 sharply to about 721 at an x CO2 of 0.985 and CO2 permeance of 5.67 × 10−7 mol m−2 s−1 Pa−1, and then deceases with a further increase in x CO2 . Similar trends in CO2 and H2 permeances, as well as α(CO2/H2) vs. x CO2 are found for the PMOF-5 membrane, but the PMOF-5 membrane shows lower H2 permeance and higher CO2 permeance in the x CO2 range of 0.9–1.0, leading to its significantly enhanced α(CO2/H2) values. α(CO2/H2) of the PMOF-5 membrane first increases tardily with increasing x CO2 , ascends drastically when x CO2 is

PMOF-5 membranes. The images difference, especially the inset images, of MOF-5 and PMOF-5 in Fig. 1 is due to the different observiation spots and random MOF-5 crystal distribution. XRD patterns of both MOF-5 and PMOF-5 membranes (Fig. 2a) and powders (Fig. 2b) show typical peaks for randomly oriented MOF-5 membranes (or powder) and match well with each other. For both powders and membranes, the CO2 posttreatment did not change the phase structure of MOF-5 and only showed a minor effect on the crystallinity. However, the single gas (CO2 or H2) permeation tests at 298 K and a feed pressure of 5 atm show that the CO2 and H2 permeances increase after CO2 post-treatment from 5.5 × 10−7 and 1.96 × 10−6 mol m−2 s−1 Pa−1 to 9.2 × 10−7 and 2.94 × 10−6 mol m−2 s−1 Pa−1, respectively. CO2/H2 gas mixture permeation/separation tests were performed with MOF-5 and PMOF-5 membranes at various feed compositions,

(b) 6000 MOF-5 PMOF-5

10 4500 Separation factor

8

-2

-1

-1

Permeance (10 mol m s Pa )

(a)

6

CO2, MOF-5

4

H2, MOF-5

-7

CO2, PMOF-5 H2, PMOF-5

2 0 0.90

3000

1500

0 0.92

0.94

0.96

0.98

1.00

0.90

0.92

0.94

0.96

0.98

1.00

CO2 mole fraction at the feed side

CO2 mole fraction at the feed side

Fig. 3. Performance comparison of MOF-5 and PMOF-5 membranes for CO2/H2 gas mixture separation at 298 K and a feed pressure of 5 atm: (a) permeance and (b) separation factor.

99

Journal of Membrane Science 555 (2018) 97–104

Z. Rui et al.

(b) 6000 o

10

25 C o

4500 25 C,

o

50 C,

o

100 C

H2:

25 C,

o

50 C,

o

100 C

o

o

-7

6

CO2:

Separation factor

8

-2

-1

-1

Permeance (10 mol m s Pa )

(a)

4

50 C o

100 C 3000

1500

2 0

0 0.90

0.92

0.94

0.96

0.98

1.00

0.90

0.92

0.94

0.96

0.98

1.00

CO2 mole fraction at the feed side

CO2 mole fraction at the feed side

Fig. 4. Performance of the PMOF-5 membrane for CO2/H2 gas mixture separation at a feed pressure of 5 atm and various temperatures.

membranes obtained in this work are above the upper bounds. Both the CO2 permeability and CO2/H2 separation factor of MOF-5 membrane increase to some extent after the CO2 post-treatment (PMOF-5), which breaks the permeability-selectivity trade-off upper bound existing in the current CO2-selective membranes. Apparently, the performance of PMOF-5 membranes is significantly better than state of the art polymeric membranes and typical zeolite membranes in spite of the difference in experimental conditions. More interestingly, even if performed at a high temperature of 373 K, both the separation factor and permeability of the PMOF-5 membrane are higher than those of typical polymer and zeolite membranes, while the CO2/H2 selectivity of most polymeric membranes and zeolite membranes decreases substantially with increasing temperature [4]. The CO2 adsorption-driven separation mechanism of the MOF-5 membranes for CO2/H2, CO2/N2 and CO2/CH4 separation has been demonstrated previously [3,19]. In the case of CO2/H2 separation in this work, the blocking effect of adsorbed CO2 on the H2 flow increases with increasing CO2 feed partial pressures and the sharp separation is achieved upon reaching the critical value. This situation corresponds to the high CO2 feed concentration case in Fig. 3, i.e., x CO2 = 98.5% for MOF-5 and 98.2% for PMOF-5. However, it can be expected that such a separation property can also be achieved at a lower CO2 concentration and higher CO2 feed pressure. The CO2 mole fraction (or feed partial pressure) required for the maxima permselectivity of MOF-5 for the CO2/H2 gas mixture (98.5%) is higher than that observed for CO2/CH4 (81.5%) and CO2/N2 (87.4%) under the same total transmembrane pressure and temperature reported in our previous work [3] because of the smaller kinetic diameter of H2 (~ 0.29 nm) in comparison with N2 (~ 0.36 nm) and CH4 (~ 0.38 nm). After reaching the maxima separation efficiency, the induced H2 permeation by the multicomponent effects of CO2 permeation [39] and its improved concentration gradient lead to moderate decrease in CO2/H2 permselectivity with further increase in CO2 feed partial pressures.

higher than 0.96, and reaches the maximum of 5781 with a CO2 permeance of ~ 9.38 × 10−7 mol m−2 s−1 Pa−1 at x CO2 = 0.982. These separation properties were measured for two high quality samples for each type of membranes, and exhibited less than 10% difference in CO2 permeance and α(CO2/H2) for both MOF-5 and PMOF-5 membranes. Fig. 4 shows that the CO2 permeance and α(CO2/H2) decrease slightly with increasing permeation temperature from 25 °C to 50 °C in the PMOF-5 membrane. Even at 100 °C, the sharp separation properties of the PMOF-5 membrane are observed. The maxima α(CO2/H2) values of PMOF-5 membrane are 5781 at 25 °C and x CO2 = 0.982, 5357 at 50 °C and x CO2 = 0.982, and 2447 at 100 °C and x CO2 = 0.980, respectively. The upper bound of CO2/H2 separation factors versus CO2 permeabilities [1 barrer = 3.348 ˣ10−16 mol m/(m2 s Pa)] suggested by Freeman et al. [5] for polymeric membranes is used to compare the performance of MOF-5 and PMOF-5 membranes with the typical CO2selective polymer based membranes and zeolite membranes. The details of the membranes and performance test conditions are listed in Table S1~S3 (Supporting information). Although there are a finite number of CO2-selective MOF membranes reported [2,18,19], they generally show much lower CO2/H2 separation selectivity than the common polymer membranes or zeolite membranes and are not included in the comparison. As compared in Fig. 5, the most popular CO2-selective membranes are still polymeric membranes to date, although some CO2-selective zeolite membranes reported exhibit a competitive CO2/H2 separation performance. The data points for MOF-5 and PMOF-5

4

10

Upper bound Zeolite Polymer MOF-5, this work PMOF-5, this work

3

Alpha CO2/H2

10

2

10

3.2. Understanding effects of CO2 post-treatment on membrane separation properties

1

10

The effect of CO2 post-treatment on the surface properties of MOF-5 membranes and powders were characterized by FTIR and in situ DRIFTS study. Because of the extremely intense IR bands of Al2O3 support, it is difficult to check the difference in the FTIR spectra between MOF-5 and PMOF-5 membranes. Alternatively, attention is directed towards the characterization of MOF-5 and PMOF-5 powders. Fig. 2b shows no difference in the XRD patterns between MOF-5 and modified PMOF-5. MOF-5 and PMOF-5 powders give BET surface areas of 319 and 323 m2/g with an error less than 5%, respectively, and same pore volumes of 0.33 cm3/g. The lower surface area of the MOF-5

0

10 -1 10

10

0

1

2

3

4

10 10 10 10 CO2 permeability (Barrer)

10

5

10

6

Fig. 5. CO2/H2 separation factors versus CO2 permeability for MOF-5 and PMOF-5 membranes. The data for the typical polymer based membranes [5,25–35], zeolite membranes [8–10,36–38] and the upper bound line for polymer membranes suggested by Freeman et al. [5] are listed for comparison.

100

Journal of Membrane Science 555 (2018) 97–104

Z. Rui et al.

1440

PMOF-5 Kubelka Munk

1440

866

PMOF-5 2954

866

MOF-5 3607

MOF-5 4000 3600 3200 2800 2400 2000 1600 1200

1600

800

1200

800

-1 Wavelength (cm )

-1 Wavelength (cm ) Fig. 6. FTIR spectra of MOF-5 and PMOF-5 powders.

and new IR peaks characteristic of the antisymmetric stretching vibration of CO32- at 1440 cm−1, out-of-plane bending mode of CO32- at 866 cm−1, and the overtone and combination band of CO32- at 2954 cm−1 appear [44]. These changes suggest an irreversible reaction pathway of CO2 with the surface impurities hydroxyl ions on the MOF-5 surface during the post-treatment by 5 atm CO2 at 100 °C, namely 2OH-+CO2 ⇒ H2O + CO32-. Fig. 7 shows the dynamic changes in the DRIFTS spectra of MOF-5 and PMOF-5 powders as a function of pure CO2 pressure at 25 °C. The CO2 pressure was stepped upward from vacuum (referred as 0 atm CO2) to 1 atm and kept at each pressure for 10 min prior to the test. Upon exposing the powders to 0.1 atm CO2, the two strong adsorption bands around 671 and 2325 cm−1 assigned to the bending mode (ν2) and asymmetric stretching mode (ν3), and the bands around 3731 and 3696 cm−1 assigned to v3+v1 mode of adsorbed CO2 appear [45]. Because the extremely intense CO2 IR bands of ν2 (~ 671 cm−1) and ν3 (~ 2325 cm−1) modes are difficult to follow and there is an original band at 3605 cm−1 of MOF-5, which is assigned to hydroxyl ions [40] and overlaps with the CO2 combination modes of ν3 + 2ν2 (around 3600 cm−1) [45], the ν3 + ν1 mode (around 3700 cm−1) is analyzed. As compared by the change of band intensity of 3731 and 3696 cm−1 in Fig. 8, the CO2 adsorption amount increases with increasing CO2 partial pressure. At a low CO2 partial pressure (0.6 atm or

powders obtained here in comparison with those reported in the literatures [12,22–24] is attributed the different MOF-5 powder synthesis procedures. In order to reveal the effect of CO2-post treatment on the properties of the MOF-5 membrane, the MOF-5 powders employed in this work were synthesized under parallel experimental conditions for the synthesis of MOF-5 membranes. The major differences with the popular synthesis procedure reported in the literatures [22,24] include the addition of a given amount of N-ethyldiisopropylamine (EDIA) and a long solvothermal synthesis time of 4 h. Fig. 6 presents the FTIR spectra of MOF-5 and PMOF-5 powders. The samples were evacuated at 100 °C overnight to remove physisorbed CO2 and water before the characterization. Therefore, FTIR spectrum changes of MOF-5 after CO2 posttreatment can be attributed to the irreversible insertion and/or chemisorption of CO2 within its crystal structure. For MOF-5 samples, the peak around 3607 cm−1 is observed, which is assigned to the O−H stretching of hydroxyl group bonding with Zn [40,41]. The existence of OH groups in MOF-5 samples has been frequently detected in literature [42,43]. Bordiga et al. [42] proposed that the OH groups present on the external surface of MOF-5 as termini of the microcrystals or at internal defects, which served as a minority of strong H2 adsorption sites. While Hafizovic et al. [43] found the presence of Zn(OH)2 species in the MOF-5 with a low surface area. After the CO2 post-treatment, the band around 3607 cm−1 disappears,

3633

3731

0.5

Kubelka Munk

3605

3633

(a)

0.5

(b)

3696

Kubelka Munk

3696 3731

1.0 atm CO 0.6 atm CO 0.3 atm CO

3605

1.0 atm CO

0.1 atm CO

0.6 atm CO 0.3 atm CO

MOF-5

0.1 atm CO PMOF-5

4000

3200 1600 -1 Wavenumber (cm )

800

4000

3200

1600 -1

Wavenumber (cm

)

Fig. 7. Dynamic changes in the DRIFTS spectra of MOF-5 (a) and PMOF-5 powders (b) at different CO2 pressures. The scale bar is included.

101

800

Journal of Membrane Science 555 (2018) 97–104

Z. Rui et al.

70

lower after CO2 treatment. The H2 and CO2 adsorption isotherms in Fig. 9 can be well fitted by Langmuir–Freundlich (L–F) isotherm equation [47]:

-1

3696 cm , MOF-5 -1 3731 cm , MOF-5 -1 3696 cm , PMOF-5 -1 3731 cm , PMOF-5

-1

S3696 or 3731 cm Intensity

60 50

q = qm

40

20 10 0 0.2

0.4 0.6 Pressure (atm)

0.8

1.0

Fig. 8. Intensity of the bands at 3731 and 3696 cm−1 as a function of pure CO2 pressure at 25 °C at 10 min adsorption time for MOF-5 and PMOF-5 powder.

lower), there is no obvious difference in CO2 adsorbed amount over MOF-5 and PMOF-5. At low CO2 pressures, CO2 molecules are expected to interact with MOF-5 by the physical adsorption or weak chemical adsorption, and no obvious reaction between MOF-5 and CO2 is observed, which is consistent with the reported CO2 adsorption properties in MOF-5 [22]. However, at 1.0 atm CO2, the intensity of the bands corresponding to PMOF-5 is obviously higher than that of MOF-5, indicating the superior CO2 uptake of PMOF-5 at a high pressure. Because both the hydroxyl groups and zinc-oxygen cluster corner are strong binding sites for H2 [42] and CO2 adsorption [1,46], the CO2 post-treatment at high pressure (or the formation of CO32-) may affect the CO2 and H2 adsorption performance. For the adsorption-driven CO2/H2 separation in the MOF-5 membrane, the performance relates to the quality of membrane and the adsorption properties of the material. The proposed CO2/H2 separation mechanism for MOF-5 and PMOF-5 membranes above was further checked by CO2 and H2 adsorption experiments in MOF-5 and PMOF-5 powders. Fig. 9a shows a modest increase in CO2 adsorption amount with increasing CO2 partial pressure and decreasing temperature for both MOF-5 and PMOF-5 samples. There is no plateau in the adsorption isotherms in the pressure range investigated, indicating that MOF-5 and PMOF-5 can adsorb more CO2 at higher CO2 partial pressures. The CO2 post-treatment exhibits a slight promotion effect on the CO2 uptake in MOF-5 at a high CO2 pressure (higher than 0.6 atm), which is consistent with the in situ DRIFTS finding (Fig. 7). Fig. 9b compares H2 adsorption isotherms for MOF-5 and PMOF-5 samples at 273 K. As compared, H2 uptake in PMOF-5 is

y1 Pt = x1 P1 and (1 − y1) Pt = (1 − x1) P2

(4)

where y1 and x1 denote the molar fractions of component 1 in the gas phase and in the adsorbed phase, respectively. Pt is the total gas pressure. P1 and P2 are the pressure of components 1 and 2 at the same spreading pressure as that of the mixture, respectively. Adsorption selectivity in a binary is defined as,

Sads,1/2 =

x1/ x2 y1 / y2

(5)

The fitted L–F isotherm equations were then applied to perform the necessary integrations in IAST. As shown in Fig. 10a, the IAST-predicted selectivities for CO2/H2 (50/50) mixtures in PMOF-5 are higher than those in MOF-5. To determine the CO2 binding affinity of MOF-5 and PMOF-5 samples, we calculate the isosteric heats of CO2 adsorption (Q, kJ/mol) using the Clausius–Clapeyron equation [47],

∂ ln P ⎤ Q = −R ⎡ ⎢ ⎣ ∂ (1/ T ) ⎥ ⎦q

(6)

where P is the pressure, T is the temperature, q refers to the amount adsorbed, R is the universal gas constant. As compared in Fig. 10b, the Q values of PMOF-5 are lower than those of MOF-5, indicating a weaker CO2-PMOF-5 interaction. From the previous analysis, we can propose that the CO2 posttreatment leads to the formation of surface CO32-, which blocks H2

3.0

0.15

MOF-5 PMOF-5 273 K 2.5 298 K 323 K 2.0 L-F fitting

H 2 uptake (mmol/g)

CO 2 uptake (mmol/g)

(3)

Here, P is the pressure of the bulk gas at equilibrium with the adsorbed phase (kPa), q is the adsorbed amount per mass of adsorbent (mmol/g), qm is the saturation capacity (mmol/g), b is the affinity coefficient of the adsorption sites (1/kPa), and n is the measure of the deviations from an ideal homogeneous surface. The values of L-F isotherm parameters and the regression coefficients are listed in Table 1. The CO2 adsorption intensity (n ≠ 1) suggests that CO2 adsorption on MOF-5 and PMOF-5 can be described by reversible adsorption and the occurrence of multilayer adsorption. According to the comparison, it can be seen that PMOF-5 is less favorable for H2 adsorption, but shows slightly better CO2 adsorption properties. The preferential adsorption selectivity of CO2 over H2 in mixture was predicted by ideal adsorbed solution theory (IAST) from the experimental pure-gas isotherms [48]. For binary adsorption of components 1 and 2, the IAST requires

30

0.0

bP1/ n 1 + bP1/ n

1.5 1.0 0.5

MOF-5 PMOF-5 L-F fitting T = 273 K

0.10

0.05

(b)

(a) 0.0

0.00 0

20

40

60

80

100

0

PCO2 (kPa)

20

40

60

80

100

PH2 (kPa)

Fig. 9. CO2 (a) and H2 (b) adsorption isotherms on MOF-5 and PMOF-5 powders, as well as their Langmuir-Freundlich (L-F) model fitting results.

102

Journal of Membrane Science 555 (2018) 97–104

Z. Rui et al.

Table 1 Parameters for Langmuir-Freundlich isotherm equation for H2 and CO2 adsorption over MOF-5 and PMOF-5 powders. Conditions

MOF-5

H2, 273 K CO2, 273 K CO2, 298 K CO2, 323 K

PMOF-5

qm (mmol/g)

b (1/kPa)

n

R2

qm (mmol/g)

b (1/kPa)

n

R2

110.36 4.73 5.49 3.76

1.05 × 10−4 0.0087 0.0043 0.0028

1.8825 0.9501 1.0311 1.0211

0.9926 0.9999 0.9999 0.9999

161.25 4.81 5.83 5.17

5.59 × 10−6 0.0075 0.0042 0.0022

0.9544 0.9051 1.0282 1.0171

0.9994 0.9999 0.9999 0.9999

24

30

Q (kJ/mol)

Selectivity of CO 2/H 2

40

MOF-5 PMOF-5 T=273 K, CO2/H2= 50/50

20

22

20 MOF-5 PMOF-5

10 18

(a)

(b)

0 0

20

40

60

80

100

0.0

0.2

0.4

0.6

0.8

CO2 adsorption amounts (mmol/g)

Pressure (kPa)

Fig. 10. IAST-predicted selectivities for CO2/H2 (50/50) mixtures in MOF-5 and PMOF-5 powders (a), and coverage-dependent CO2 adsorption enthalpy obtained by LangmuirFreundlich (L-F) fitting (b).

adsorption sites and lowers the H2 adsorption capacity (Fig. 9b), as well as weakens the interaction between PMOF-5 and CO2 (Fig. 10b). In consideration with multilayer adsorption [22] and the important role of CO2-CO2 electrostatic interactions for CO2 adsorption in MOF-5 [46], the CO2 post-treatment shows negligible effect on the CO2 adsorption capacity and even a promotion effect at a high CO2 pressure (Fig. 9a). As a result, the CO2/H2 adsorption selectivity of MOF-5 increases after the CO2 post-treatment (Fig. 10a). The weakened CO2-PMOF-5 interaction and enhanced CO2/H2 adsorption selectivity of PMOF-5 are the major reasons for the remarkable enhancement of the CO2 adsorption driven CO2/H2 separation factor together with the slight CO2 permeance enhancement in PMOF-5 membrane at the optimized feed conditions.

annealing treatment sheds a new light on synthesis of highly perm-selective MOF membranes for gas separations. Acknowledgement The work was supported by Petroleum Research Fund, administrated by the American Chemical Society (50928-ND9) and National Science Foundation (CBET-1160084 & CBET-1511005). ZB Rui is grateful to China Scholarship Council for fellowship to support his visit to ASU. We thank Prof. Bin Mu at ASU for use of his in situ DRIFTS instrumentation. Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.memsci.2018.03.036.

4. Conclusions Extremely high separation efficiency of MOF-5 membranes for CO2 separation from dry and CO2 enriched CO2/H2 mixture based on a CO2 adsorption-driven mechanism was reported. The IRMOF-1 membrane exhibited a CO2/H2 separation factor of 721 with CO2 permeance of 5.67 × 10−7 mol m−2 s−1 Pa−1 at CO2 mole fraction of 98%, feed pressure of 5 atm and 298 K, respectively. Post-treatment of the MOF-5 membrane by on-stream CO2 atmosphere at 100 °C and 5 atm deceases H2 permeance while slightly increasing CO2 permeance, leading to the simultaneous enhancement in both CO2/H2 separation factor up to ~ 5781 and CO2 permeance up to ~9.38 × 10−7 mol m−2 s−1 Pa−1 at a CO2 mole fraction of 98% and 5 atm. The post-treated MOF-5 membrane demonstrates sharp separation properties for CO2/H2 separation and breaks the permeability and selectivity trade-off limitation, which is remarkably better than MOF-5 membranes and state of the art zeolite (or polymeric) membranes. The unusual separation behavior of the post-treated MOF-5 membrane is related to the enhanced CO2 adsorption selectivity over H2 and reduced CO2 affinity of the membrane caused by the formation of surface CO32- during the CO2 treatment. The enhanced adsorption controlled CO2 selectivity with post CO2

References [1] J. Yu, L. Xie, J. Li, Y. Ma, J.M. Seminario, P.B. Balbuena, CO2 capture and separations using MOFs: computational and experimental studies, Chem. Rev. 117 (2017) 9674–9754. [2] B.A. Al-Maythalony, O. Shekhah, R. Swaidan, Y. Belmabkhout, I. Pinnau, M. Eddaoudi, Quest for anionic MOF membranes: continuous sod-ZMOF membrane with CO2 adsorption-driven selectivity, J. Am. Chem. Soc. 137 (2015) 1754–1757. [3] Z. Rui, J.B. James, A. Kasik, Y.S. Lin, Metal-organic framework membrane process for high purity CO2 production, AIChE J. 62 (2016) 3836–3841. [4] P. Li, Z. Wang, Z. Qiao, Y. Liu, X. Cao, W. Li, J. Wang, S. Wang, Recent developments in membranes for efficient hydrogen purification, J. Membr. Sci. 495 (2015) 130–168. [5] H. Lin, E. Van Wagner, B.D. Freeman, L.G. Toy, R.P. Gupta, Plasticization-enhanced hydrogen purification using polymeric membranes, Science 311 (2006) 639–642. [6] H.B. Park, J. Kamcev, L.M. Robeson, M. Elimelech, B.D. Freeman, Maximizing the right stuff: the trade-off between membrane permeability and selectivity, Science 356 (2017) 1–10. [7] W.J. Koros, R. Mahajan, Pushing the limits on possibilities for large scale gas separation: which strategies? J. Membr. Sci. 175 (2000) 181–196. [8] M. Hong, S. Li S, J.L. Falconer, R.D. Noble, Hydrogen purification using a SAPO-34 membrane, J. Membr. Sci. 307 (2008) 277–283. [9] M. Zhou, D. Korelskiy, P. Ye, M. Grahn, J. Hedlund, A uniformly oriented MFI

103

Journal of Membrane Science 555 (2018) 97–104

Z. Rui et al.

[10] [11]

[12]

[13] [14]

[15] [16] [17]

[18] [19] [20] [21]

[22] [23]

[24] [25]

[26]

[27]

[28]

[29]

[30] I. Taniguchi, S. Duan, S. Kazama, Y. Fujioka, Facile fabrication of a novel high performance CO2 separation membrane: immobilization of poly (amidoamine) dendrimers in poly(ethylene glycol)networks, J. Membr. Sci. 322 (2008) 277–280. [31] Z. Qiao, Z. Wang, S. Yuan, J. Wang, S. Wang, Preparation and characterization of small molecular amine modified PVA membranes for CO2/H2 separation, J. Membr. Sci. 475 (2015) 290–302. [32] Y. Zhao, W.S. Winston Ho, Steric hindrance effect on amine demonstrated in solid polymer membranes for CO2 transport, J. Membr. Sci. 415–416 (2012) 132–138. [33] Y. Zhao, W.S. Winston Ho, CO2-selective membranes containing sterically hindered amines for CO2/H2 separation, Ind. Eng. Chem. Res. 52 (2013) 8774–8782. [34] Y. Li, T.-S. Chung, Molecular-level mixed matrix membranes comprising Pebax and POSS for hydrogen purification via preferential CO2 removal, Int. J. Hydrog. Energy 35 (2010) 10560–10568. [35] C.H. Lau, S. Liu, D.R. Paul, J. Xia, Y.-C. Jean, H. Chen, L. Shao, T.-S. Chung, Silica nanohybrid membranes with high CO2 affinity for green hydrogen purification, Adv. Energy Mater. 1 (2011) 634–642. [36] J. Lindmark, J. Hedlund, Carbon dioxide removal from synthesis gas using MFI membranes, J. Membr. Sci. 360 (2010) 284–291. [37] L. Sandström, E. Sjöberg, J. Hedlund, Very high flux MFI membrane for CO2 separation, J. Membr. Sci. 380 (2011) 232–240. [38] W.J.W. Bakker, F. Kapteijn, J. Poppe, J.A. Moulijn, Permeation characteristics of a metal-supported silicalite-1 zeolite membrane, J. Membr. Sci. 117 (1996) 57–78. [39] S. Keskin, D.S. Sholl, Assessment of a metal-organic framework membrane for gas separations using atomically detailed calculations: CO2, CH4, N2, H2 mixtures in MOF-5, Ind. Eng. Chem. Res. 48 (2009) 914–922. [40] P. Liao, H. Chen, D. Zhou, S. Liu, C. He, Z. Rui, H. Ji, J. Zhang, X. Chen, Monodentate hydroxide as a super strong yet reversible active site for CO2 capture from high humidity flue gas, Energy Environ. Sci. 8 (2015) 1011–1016. [41] Y. Ming, J. Purewal, J. Yang, C. Xu, R. Soltis, J. Warner, M. Veenstra, M. Gaab, U. Müller, D.J. Siegel, Kinetic stability of MOF-5 in humid environments: impact of powder densification, humidity level, and exposure time, Langmuir 31 (2005) 4988–4995. [42] S. Bordiga, J.G. Vitillo, G. Ricchiardi, L. Regli, D. Cocina, A. Zecchina, B. Arstad, M. Bjørgen, J. Hafizovic, K.P. Lillerud, Interaction of hydrogen with MOF-5, J. Phys. Chem. B 109 (2005) 18237–18242. [43] J. Hafizovic, M. Bjørgen, U. Olsbye, P.D.C. Dietzel, S. Bordiga, C. Prestipino, C. Lamberti, K.P. Lillerud, The inconsistency in adsorption properties and powder XRD data of MOF-5 is rationalized by framework interpenetration and the presence of organic and inorganic species in the nanocavities, J. Am. Chem. Soc. 129 (2007) 3612–3620. [44] R.L. Frost, W.N. Martens, D.L. Wain, M.C. Hales, Infrared and infrared emission spectroscopy of the zinc carbonate mineral smithsonite, Spectrochim. Acta Part A 70 (2008) 1120–1126. [45] Y. Hu, Z. Liu, J. Xu, Y. Huang, Y. Song, Evidence of pressure enhanced CO2 storage in ZIF-8 probed by FTIR spectroscopy, J. Am. Chem. Soc. 135 (2013) 9287–9290. [46] K.S. Walton, A.R. Millward, D. Dubbeldam, H. Frost, J.J. Low, O.M. Yaghi, R.Q. Snurr, Understanding inflections and steps in carbon dioxide adsorption isotherms in metal-organic frameworks, J. Am. Chem. Soc. 130 (2008) 406–407. [47] R. Li, M. Li, X. Zhou, D. Li, M. O’Keeffe, A highly stable MOF with a rod SBU and a tetracarboxylate linker: unusual topology and CO2 adsorption behaviour under ambient conditions, Chem. Commun. 50 (2014) 4047–4049. [48] A.L. Myers, J.M. Prausnitz, Thermodynamics of mixed gas adsorption, AIChE J. 11 (1965) 121–130.

membrane for improved CO2 separation, Angew. Chem. Int. Ed. 53 (2014) 3492–3495. J. Lindmark, J. Hedlund, Modification of MFI membranes with amine groups for enhanced CO2 selectivity, J. Mater. Chem. 20 (2010) 2219–2225. H. Guo, G. Zhu, H. Li, X. Zou, X. Yin, W. Yang, S. Qiu, R. Xu, Hierarchical growth of large-scale ordered zeolite silicalite-1 membranes with high permeability and selectivity for recycling CO2, Angew. Chem. Int. Ed. 45 (2006) 7053–7056. H.L. Li, M. Eddaoudi, M. O’Keeffe, O.M. Yaghi, Design and synthesis of an exceptionally stable and highly porous metal-organic framework, Nature 402 (1999) 276–279. S.R. Venna, M.A. Carreon, Metal organic framework membranes for carbon dioxide separation, Chem. Eng. Sci. 124 (2015) 3–19. W.B. Li, Y.F. Zhang, Q.B. Li, G.L. Zhang, Metal-organic framework composite membranes: synthesis and separation applications, Chem. Eng. Sci. 135 (2015) 232–257. M.S. Denny, J.C. Moreton, L. Benz, S.M. Cohen, Metal–organic frameworks for membrane-based separations, Nat. Rev. Mater. 1 (2016) 1–17. W.J. Koros, C. Zhang, Materials for next-generation molecularly selective synthetic membranes, Nat. Mater. 16 (2017) 289–297. X. Wang, C. Chi, K. Zhang, Y. Qian, K.M. Gupta, Z. Kang, J. Jiang, D. Zhao, Reversed thermo-switchable molecular sieving membranes composed of two-dimensional metal-organic nanosheets for gas separation, Nat. Commun. 8 (2017) 14460. S. Takamizawa, Y. Takasaki, R. Miyake, Single crystal membrane for anisotropic and efficient gas permeation, J. Am. Chem. Soc. 132 (2010) 2862–2863. Z. Zhao, X. Ma, A. Kasik, Z. Li, Y.S. Lin, Gas separation properties of metal organic framework (MOF-5) membranes, Ind. Eng. Chem. Res. 52 (2013) 1102–1108. S.M. Cohen, Postsynthetic methods for the functionalization of metal–organic frameworks, Chem. Rev. 112 (2012) 970–1000. A. Huang, N. Wang, C. Kong, J. Caro, Organosilica-functionalized zeolitic imidazolate framework ZIF-90 membrane with high gas-separation performance, Angew. Chem. 124 (2012) 10703–10707. Z. Zhao, Z. Li, Y.S. Lin, Adsorption and diffusion of carbon dioxide on metal-organic framework (MOF-5), Ind. Eng. Chem. Res. 48 (2009) 10015–10020. S. Gadipelli, Z. Guo, Postsynthesis annealing of MOF-5 remarkably enhances the framework structural stability and CO2 uptake, Chem. Mater. 26 (2014) 6333–6338. Z.X. Zhao, X.L. Ma, Z. Li, Y.S. Lin, Synthesis, characterization and gas transport properties of MOF-5membranes, J. Membr. Sci. 382 (2011) 82–90. H. Chen, Y. Xiao, T.-S. Chung, Synthesis and characterization of poly (ethylene oxide) containing copolyimides for hydrogen purification, Polymer 51 (2010) 4077–4086. W. Yave, A. Car, J. Wind, K.V. Peinemann, Nanometric thin film membranes manufactured on square meter scale: ultra-thin films for CO2 capture, Nanotechnology 21 (2010) 395301. S. Li, Z. Wang, C. Zhang, M. Wang, F. Yuan, J. Wang, S. Wang, Interfacially polymerized thin film composite membranes containing ethylene oxide groups for CO2 separation, J. Membr. Sci. 436 (2013) 121–131. H.Z. Chen, Z. Thong, P. Li, T.-S. Chung, High performance composite hollow fiber membranes for CO2/H2 and CO2/N2 separation, Int. J. Hydrog. Energy 39 (2014) 5043–5053. H.Z. Chen, P. Li, T.-S. Chung, PVDF/ionic liquid polymer blends with superior separation performance for removing CO2 from hydrogen and flue gas, Int. J. Hydrog. Energy 37 (2012) 11796–11804.

104