Ultrason® mixed-matrix membranes for selective CO2 capture from humid post combustion flue gas

Accepted Manuscript Water-stable ZIF-300/Ultrason® mixed-matrix membranes for selective CO2 capture from humid post combustion flue gas

Muhammad Sarfraz, M. Ba-Shammakh PII: DOI: Reference:

S1004-9541(17)30954-0 doi:10.1016/j.cjche.2017.11.007 CJCHE 969

To appear in: Received date: Revised date: Accepted date:

25 July 2017 1 November 2017 10 November 2017

Please cite this article as: Muhammad Sarfraz, M. Ba-Shammakh , Water-stable ZIF-300/ Ultrason® mixed-matrix membranes for selective CO2 capture from humid post combustion flue gas. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Cjche(2017), doi:10.1016/j.cjche.2017.11.007

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Separation Science and Engineering

Water-stable ZIF-300/Ultrason® mixed-matrix membranes for selective CO2 capture from humid post combustion flue gas

IP

T

Muhammad Sarfraz1,* and M. Ba-Shammakh2 1 Department of Polymer and Process Engineering, University of Engineering and Technology, Lahore-54890, Pakistan 2 Department of Chemical Engineering, King Fahd University of Petroleum and Minerals, Dhahran-31261, KSA * Corresponding Author: [email protected]

PT

ED

M

AN

US

CR

Abstract Water stable mixed-matrix membranes (MMMs) were developed to help control the global warming by capturing and sequestrating carbon dioxide (CO2) from post combustion flue gas originated from burning of fossil fuels. MMMs of different compositions were prepared by doping glassy polymer Ultrason® S 6010 (US) with nanocrystals of zeolitic imidazolate frameworks (ZIF-300) in varying degrees. Solution-casting technique was used to fabricate various MMMs to optimize their CO2 capturing performance from both dry and wet gases. The prepared composite membranes indicated enhanced filler-polymer interfacial adhesion, consistent distribution of nanofiller, and thermally established matrix configuration. CO2 permeability of the membranes was enhanced as demonstrated by gas sorption and permeation experiments performed under both dry and wet conditions. As compared to neat Ultrason ® membrane, CO2 permeability of the composite membrane doped with 40 wt % ZIF-300 nanocrystals was increased by four times without disturbing CO2/N2 ideal selectivity. In contrast to majority of previously reported membranes, key features of the fabricated MMMs include their structural stability under humid conditions coupled with better and unaffected gas separation performance.

CE

Key Words Hydrophobic MMMs, ZIF-300, gas permeation, CO2 capture, permselectivity

AC

1. Introduction Global warming issues can be resolved to a great extent by controlling CO2 emissions by its separation from post combustion flue gases, thus protecting the world environment [1]. CO2 capture or sequestration from a carbon-enriched gas mixture is one of the most cost-effectively feasible strategies to control carbon releases [2]. Amongst other typical carbon capture operations, gas separation process using polymer-based mixed-matrix membranes has gained significant importance due to its low energy requirement, high efficiency, easiness of scale up, simple design, uncomplicated function, economical operating costs and capital and environmental kindliness [3]. The key parameters of a superior quality gas separation membrane include improved permselectivity and separation factor, good mechanical strength, improved chemical and thermal stability, and good operational stability [4]. Due to its inherent structural constraints (chain mobility and inter-chain spacing), a glassy polymer membrane being highly permeable is generally less selective [5]; it requires considerable improvements for practical applications. A number of glassy polymers (such as poly(vinyl acetate), polydimethylsiloxane, polyimide, poly-

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

AN

US

CR

IP

T

(1,4-phenylene ether-ether-sulfone), polysulfone, Ultem®, Matrimid®) and inorganic nanofillers (like zeolites, carbon nanotubes [6], structured mesoporous and nonporous silica [7], microporous metal organic frameworks (MOFs) [8], and carbon molecular sieves [9]) have been used to fabricate hybrid membranes providing improved separation performance in comparison to their bare-polymer counterparts. The major advantages of developing MMMs by doping thermoplastic polymer matrix with microporous nanocrystals of MOF materials comprise the capability of coupling the easy processability and casting, improved chemical and mechanical performance of polymers with the superior gas separation effectiveness, variable pore sizes, adjustable surface functionalities, and large surface areas of microporous nanomaterials [10]. Owing to their high CO2 permeability and CO2/N2 permselectivity values hybrid membranes fabricated from commercially available polysulfone and its different grades have acquired considerable research attention for carbon capture. Selective chemical functionalization or formulation of a composite membrane by integrating microporous nanofillers into base polymer matrix are valuable strategies to further augment gas permeability through a flexible polymer. Incorporation of HKUST-1 contents into polysulfone to obtain HKUST-1/PSF hybrid membranes leads to improve both CO2 permeability and CO2/N2 selectivity [11]. Fabrication of composite membranes by doping Udel® (a commercial grade of polysulfone) with varying contents of mesoporous silica spheres helped to improve CO2 separation performance [12]. MOFs microporous nanocrystals are significant materials to formulate proficient mixed membranes for gas separation owing to their adjustable nano-sized dimensions, high surface areas, superior wetting features, improved chemical and thermal stability, and unique surface functionality [10]. Since the hydrothermally stable crystals of ZIF-300 selectively detain CO2 gas from dry and wet CO2/N2 gas mixtures [13], their insertion into glassy polymers is anticipated to enhance separation efficiency of hybrid membranes under both dry and wet environments. Hydrophobic organic linkers combined with zinc salt result into microporous isoreticularstructured ZIF-300 crystals rendering controlled pore size of few nanometers. Contrary to several MOF crystalline materials [14-16], ZIF-300 crystals do not disintegrate at elevated temperatures and effectively detain CO2 from CO2/N2/H2O mixture subjected to genuine post combustion conditions. The volumetric composition of an ordinary flue gas stream consists of 75% N2, 15% CO2, 6% H2O and 4% detectable gases [15]. Major challenges regarding carbon capture from post combustion flue gas can be addressed by using ZIF-300 material due to its improved CO2 adsorption capacity, high CO2/N2 selectivity and low regeneration enthalpy subjected to moist circumstances [13]. Expecting the improvement in gas separation performance (both under dry and wet conditions) by inserting them into glassy polymers, ZIF-300 nanocrystals have been incorporated into glassy Ultrason® (a commercial grade of polysulfone) matrix to fabricate ZIF-300/US mixed-matrix membranes for effective carbon capture. The central focus of the present work is to investigate the dependence of CO2 permeability and CO2/N2 permselectivity on ZIF-300 contents, the effect of moisture contents on CO2 separation efficiency of ZIF-300/US MMMs, and the prediction and comparison of experimentally-acquired gas permeation data with two- and three-phase permeation models. 2. Experimental and Data Validation 2.1 Materials Synthesizing ingredients of ZIF-300 i.e. zinc nitrate hexahydrate, 2-methylimidazole and 5(6)bromobenzimidazole were purchased from Merck Chemical Company. Methanol, chloroform, 2

ACCEPTED MANUSCRIPT and N,N-dimethylformamide (DMF) solvents were obtained from Aldrich Chemical Company. Commercial-grade Ultrason® S 6010, having density 1.25 g·cm-3 and LS average molecular weight ~35000, was obtained from Sigma Aldrich. The glassy polymer and all reagent-grade chemicals were utilized in the obtained form without any processing. He, N2 and CO2 gases used for gas sorption and permeation experiments were highly pure.

US

CR

IP

T

2.2 Formulation of ZIF-300 nanocrystals ZIF-300 nanocrystals were prepared via cold synthesis process by dissolving zinc nitrate hexahydrate (67.8 mg), 2-methylimidazole (23.4 mg) and 5(6)-bromobenzimidazole (56.4 mg) in a 10 mL mixture of distilled water (0.5 ml) and DMF (9.5 ml) in a 20-ml vial and sonicating for 10 minutes. The prepared solution was then gently stirred on a hot plate for 70 h by maintaining the temperature at 50 °C to obtain brown slurry of ZIF-300 nanocrystals. Centrifugation was employed to separate suspended nanocrystals from mother liquor. The resulting nanoctystals were purified by daily washing them with 7 ml fresh DMF for five times followed by solvent exchange with anhydrous methanol thrice a day at room temperature for another three days. Spent methanol was decanted followed by washing the nanocrystals thrice with chloroform.

CE

PT

ED

M

AN

2.3 Preparation of ZIF-300/US MMMs Prior to fabricate composite membranes of varying compositions all the constituents (US and ZIF-300) were degassed at 100 oC under vacuum for 20 h to get rid of adsorbed moisture and/or gases. Neat Ultrason® membrane was fabricated by dispersing 1g polymer pellets in 7 ml DMF solvent, followed by vigorously stirring at room temperature for 20 h until a thick viscous solution was obtained. The hybrid membranes doped with varying amounts of ZIF-300 nanocrystals were fabricated by incorporating 1g base polymer in 5 ml DMF followed by robust stirring at room temperature for 20 h to get a thick solution. Specified amount of ZIF-300 nanocrystals was redistributed in1 ml DMF solvent, sonicated for 10 min, and uniformly dispersed solution was obtained. Both the viscous solutions were combined collectively into a beaker and stirred for 12 h until a homogenized thick solution was acquired. After adjusting their gate height, the viscous solutions were distributed on clean glass plates and knife cast to get thin membranes. The entrapped solvent within the membranes was allowed to slowly evaporate at room temperature for 20 h. The membranes were peeled off from the glass plates and dried at 80 o C (12 h), 100oC (20 h) and 160 oC (20 h) to make them bone-dry.

AC

2.4 Membranes Characterization Methods The characterization techniques used to measure various properties of ZIF-300 nanocrystals and bare polysulfone and composite membranes include X-ray diffraction, scanning electron microscopy, thermal gravimetric analysis and gas sorption analysis. The nanofiller’s fractional volume (Ф D) incorporated into the hybrid membranes can be determined by the following equation: ФD =

𝑚D 𝜌D 𝑚D 𝑚C + 𝜌D 𝜌C

× 100

(1)

here m and ρ respectively symbolize mass and density of continuous polymer phase C (Ultrason®) and dispersed filler’s phase D (ZIF-300 nanocrystals).

3

ACCEPTED MANUSCRIPT

CR

IP

T

XRD characterization of the hybrid membranes was accomplished to confirm whether the nanofillers crystalline structure is not damaged even after their addition into the polymer matrix. The membrane small specimens were put on silicon substrate placed in a sample holder. PowderXRD radiation diagrams of all the prepared membranes were recorded by Bruker D8 X-ray Diffractometer via Cu Kα radiation (λ = 0.15406 nm) functioned at 45 mA current, 40 kV voltage and step size increase of 0.02o in 2θ. Hitachi S-4300SE/N SEM instrument was employed to scan the membrane surface to get relevant micrographs aiming to check their morphological configuration and matrix-filler interfacial aspects. Prior to testing, the membrane samples were prepared by cryofracturing them in liquid nitrogen followed by coating their external surfaces with a delicate golden film to circumvent electrons charging. The SEM device was run at 20 kV to take fine quality images. Thermal gravimetric analysis was employed to determine thermal constancy and other associated attributes of membrane materials via TGA/SDTA 851 (Mettler Toledo) system operated in air. Testing specimen was heated from room temperature to 700 oC by maintaining the heating rate at

US

10 oC·min-1. Differential scanning calorimetry (DSC) was carried out via Netzsch DSC 200F3 Calorimeter to assess glass transition temperature (Tg) of the membranes. The equipment was operated in nitrogen atmosphere being flown at a rate of 50 ml·min-1 and heated from 40 to 200

PT

ED

M

AN

°C while maintain a heating rate of 5°C·min-1. Quantachrome Autosorb iQ Gas Sorption Analyzer was used to get N2 and CO2 adsorption isotherms at different temperatures of 77 and 298 K. After cutting them into minute bits, the membrane specimens were degassed, subjected to vacuum (<0.1 Pa), at 100 oC for 5 h. The isotherms obtained at 77 K were analyzed to obtain different microporous properties (such as N2 and CO2 gas uptakes, microporous volume, specific Langmuir and BET surface areas etc.) of the membrane materials. The sorption measurements evaluated at a temperature of 298 K in the pressure range of 0.1 to 1×105 Pa were employed to assess physisorption data of N2 and CO2 gases.

AC

CE

2.5 Gas Permeation Measurements Single gas permeation cell operated in constant-volume (variable-pressure) mode at a temperature of 298 K [17-18] was utilized to evaluate membrane transport properties (permeability) and separation performance (ideal selectivity) for N2 and CO2 gases. Average thickness of the membranes, in the range of 70-120 µm, was measured by a digital micrometer and the membrane sample was placed in the permeation cell to perform permeation test. The permeate-side line of the cell was evacuated by keeping the upstream valve closed and downstream valve connecting to a vacuum pump. Once the vacuum was generated on the permeate side, the valve between vacuum pump and permeate-side line was shut down and feedside valve opened to uphold a constant pressure (e.g., 104 Pa) on feed-side for a definite time period (i.e., 2 h) to register the permeation measurements. The upstream pressure was next intermittently raised followed by taking the reading after 1 h of stabilization for each step. To ensure their perfection and certainty, at least three replicas were fabricated and assessed equivalent to every membrane specimen. Gas permeability of the membrane can be assessed using the following relation: 4

ACCEPTED MANUSCRIPT

𝑃𝑖 =

22414 𝐴

𝑉

𝑙 d𝑃𝑖

× 𝑅𝑇 × Δ𝑃

𝑖

(2)

d𝑡

𝑙2 6𝜃

[1 +

6𝐾 𝑦 2 𝑦3

{

2

𝑉

+ 𝑦 − (1 + 𝑦) ln(1 + 𝑦)} ( d )]

CR

𝐷=

IP

T

where Pi, A, V, R, T, l, ∆Pi and ΔPi/dt denote gas permeability (Barrer, 1 Barrer = 10-9 mol·m-2· s-1·Pa-1), membrane effective area (cm2), downstream chamber volume (cm3), universal gas constant (6236.56 cm3·cmHg·mol-1·K-1, 1cmHg=13.3322Pa), absolute temperature (K), membrane thickness (cm), pressure difference across the membrane (psi, 1 psi=6894.76Pa), and gas permeation rate (psi·s-1) of component i respectively. Diffusion coefficient (D) of the membrane was determined via diffusivity (D) vs. time (θ) lag relationship as suggested by Paul and Kemp [19]:

𝑉p

(3)

US

here Vp and Vd denote volume fractions of polymer and filler phases respectively; y and K represent adsorption parameters to be calculated from Langmuir adsorption isotherm. The solubility coefficient (S) of the membrane was calculated via following equation: 𝑃

𝑆=𝐷

AN

(4)

M

Ideal selectivity (αij) of gas i over j was computed from the eq. given below: 𝑃

𝐷

𝑆

𝛼𝑖𝑗 = 𝑃 𝑖 = (𝐷 𝑖 ) (𝑆 𝑖 ) 𝑗

𝑗

(5)

ED

𝑗

Here (Si/Sj) and (Di/Dj) denote respectively the solubility- and diffusion-based selectivity terms.

AC

CE

PT

2.6 Predicting Permeability of ZIF-300/US MMMs Gas permeation through a membrane can be estimated via different theoretical models [20] based on ideal (two-phase system) and non-ideal (three-phase system) morphologies of composite membranes. A two-phase permeation model based on perfect morphology of MMMs can be characterized by a continuous polymer matrix phase and a homogeneously distributed nanoparticles phase possessing a faultless, defect-free and non-deformable polymer-filler interface. A non-ideal morphology representing imperfections, defects and faults at polymerfiller interface leads to a three-phase system which takes into consideration the polymer-filler interface along with matrix and filler phases. Gas permeability of a composite membrane relative to its bare polymer counterpart (Pr) can be predicted via following expressions based on two-phase permeation models. 𝜆

Maxwell:

𝑃r =

dm

𝜆dm −1

1 − 𝛷 (𝜆

dm

Bruggeman:

1 3

𝑃r (

−1

1 + 2𝛷 (𝜆dm +2) ) +2

𝜆dm − 1 ) = (1 − 𝛷)−1 𝜆dm − 𝑃r

(6)

(7) 5

ACCEPTED MANUSCRIPT

Singh:

𝑃r = 1 + 3.74 (

𝜆dm − 1 2/3 )𝛷 𝜆dm + 2

(8)

where 𝑃eff Effective permeability of MMM = 𝑃m Permeability of polymer matrix 𝑃d Permeability of dispersedphase = 𝑃m Permeability of polymer matrix

T

𝜆dm =

CR

1 − 𝛷m 𝛹 =1+( )𝛷 2 𝛷m

IP

𝑃r =

US

𝛷m = volume fraction of fillers at maximum packing = 0.64 Φ = volume fraction of fillers

𝑃r =

1−

(𝛽+2𝛾) 𝛷𝛹(𝛽−𝛾)

(9)

(𝛽+2𝛾)

ED

Modified Felske:

2𝛷(𝛽−𝛾)

M

1+

AN

The three-phase permeation model used to anticipate the relative permeability of a gas through a composite membrane can be expressed as follows:

where

CE

PT

𝛽 = (2 + 𝛿 3 )𝜆dm − 2(1 − 𝛿 3 )𝜆im

AC

𝜆di =

𝛾 = 1 + 2𝛿 3 − (1 − 𝛿 3 )𝜆di

𝑃d Permeability of dispersedphase = 𝑃i Permeability of interphase

𝜆im =

𝑃i Permeability of interphase = 𝑃m Permeability of matrixphase 𝜆dm = 𝜆di 𝜆im =

𝛽 𝛾

𝛿 = ratio of interphase to particle radii 3. Results and discussion The end results of different characterization techniques i.e., XRD, SEM, TGA, gas sorption and gas permeation are briefly described here. 6

ACCEPTED MANUSCRIPT

Bare Ultrason 10% ZIF-300/US 20% ZIF-300/US 30% ZIF-300/US 40% ZIF-300/US

10

Activated ZIF-300

20

30

2θ /

40

50

(o)

AC

0

CE

PT

ED

Intensity

M

AN

US

CR

IP

T

3.1 Powder X-ray Diffraction It is a valuable technique to examine the nanofillers impact on polymer chains arrangement in MMMs and to substantiate the continuation of ordered structure of crystalline materials after incorporating them into polymers. In order to ascertain actuality of crystalline of CNTs and ZIF300 nanocrystals in MMMs, XRD quantifications were recorded in the 2θ range of 2o-50o (see Fig. 1). Intensity peaks arising at 2θ locations of 5.3o, 6.5o, 11.3o, 17.2o and 22.9º in XRD schemes of composite membranes confirm the dispersion of crystalline structure of isoreticular ZIF-300 nanocrystals [39] within Ultrason® matrix. The preservation of crystalline structure of these nanocrystals even after their incorporation into the polymer matrix is well supported by XRD patterns of MMMs. Also the peak heights at particular 2θ levels correspond to respective nanofillers loadings in MMMs. In addition, the centralization of a specific broad peak at 2θ position of 17.2o is associated with base polymer membrane. Also the minor reallocation of the wide peak of unfilled Ultrason® from a 2θ position of 17.2° (d-spacing = 0.521 nm) to 17.3° (dspacing = 0.517 nm), indicating the reduction in polymer inter-chain distance, can be ascribed to intense filler-polymer interactions. Decrease in inter-chain dimensional attitude endorses CO2/N2 permselectivity enhancement due to the fact of size exclusion.

Fig. 1. XRD patterns of pure ZIF-300, pure Ultrason®, and ZIF-300/US MMMs containing 10%, 20%, 30% and 40 % by mass of ZIF-300 3.2 Morphology of ZIF-300/US MMMs The characteristic features of polymer matrix and incorporated nanofillers mainly dictate the distinctive interior microstructural features and morphology of fabricated MMMs. In order to explore their morphological features, nanofillers distribution within polymer substance and polymer-filler interfacial adhesion, cross sectional photomicrographs of composite membranes were probed via SEM. The internal surfaces of all MMMs doped with varying amounts of nanofillers depicted continuous appearances, virtually independent of interfacial spaces. Cross sectional outlook of these composite membranes demonstrated faultless interrelated morphology, 7

ACCEPTED MANUSCRIPT

B

C

AN

US

A

CR

IP

T

effective adhesion at polymer-filler interface and uniform dispersal of nanofillers in polymer matrix. Hybrid membranes doped with varying loadings of nanofiller exhibit uninterrupted phases for almost all MMMS (Fig. 2) due to homogeneous distribution of tiny nanocrystals of ZIF-300. The phase continuity at filler-matrix boundary leads to voids-free interface in all MMMs. Nanofiller distribution in MMMs filled with low loadings (10 wt%) of ZIF-300 nanocrystals was sparse (Fig.2B). Hybrid membranes doped with moderate contents (20wt%-30 wt%) of ZIF-300 nanocrystals manifested consistent dispersion of nanofiller in polymer matrix (see Fig.2C, Fig.2D). High loadings (40 wt%) of ZIF-300 nanoparticles depict signs of slightly rough surface and marginal aggregation (Fig.2E). Specific spatial arrangement of chabazite-type ZIF-300 nanocrystals within polymer matrix lead to smooth, voids-free and crack-less morphology corresponding to their defined loadings.

1 μm

E

1 μm

F

PT

ED

M

D

1 μm

1 μm

1 μm

1 μm

CE

Fig. 2. Scanning electron micrographs of pure Ultrason® (A), ZIF-300/US MMMs containing 10% (B), 20% (C), 30% (D), 40 % (E) by mass of ZIF-300, and pure ZIF-300 (F).

AC

3.3 Thermal Gravimetric Analysis The effect of incorporating ZIF-300 nanocrystals into neat Ultrason® matrix, determined in terms of thermal stability, phase transitions, physical and chemical performance etc., by carrying out thermal gravimetric analysis (TGA) coupled with derivative thermal gravimetric (DTG) analysis [collectively called (TGA-DTG) analyses] of bare Ultrason® and composite membranes in the temperature range of 40-700 oC. Thermal phase transitions of fabricated membranes were depicted by TGA decomposition profiles (see Fig. 3A) in terms of specimen mass loss taking place in two steps. The two-stage mass loss takes place due to desolvation and pyrolysis processes in the temperature ranges of 100-200 oC and 510-640 oC respectively. The former occurs on account of liberation of volatile solvent molecules (DMF, water, CCl4 etc.) from the membrane pores. The latter owes to the degradation of Ultrason® structure into its constituting entities (C, benzene, SO2, toluene, phenol, 8

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

AN

US

CR

IP

T

xylene, styrene etc.) accompanied by the degeneration of organic ligands of imidazole frameworks (5(6)-bromobenzimidazole and 2-methylimidazole). Minimal mass contents of the nanofillers added to hybrid membranes were demonstrated by the ash leftover in the pan on completion of the experiment. Td5% and Td10% values, defined as the temperatures at which a material sample loses its 5 and 10 percent mass respectively, facilitates to assess its thermal constancy [21]. These values are controlled by the loadings of constituting nanofillers of composite membranes and normally fall in the temperature range of 165-175 °C and 210-450 °C respectively. Differential mass loss (DTG) curves and glass transition temperature (Tg) of the polymer specimen impart essential information on extent of stiffness of polymer chains in composite membranes. Improvement in these values (especially Tg) by increasing nanofillers loadings indicate enhanced MMMs stiffness owing to restricted polymer chains movement which is a direct consequence of polymer chain-to-nanofiller interactions. The 1stand 2nd DTG peaks in Fig. 3B render necessary data about pyrolysis rates of the specimen. Values of various important thermal properties of composite membranes are outlined in Table 1.

9

ACCEPTED MANUSCRIPT A

100

60

20

IP

40

T

Bare Ultrason 10% ZIF-300/US 20% ZIF-300/US 30% ZIF-300/US 40% ZIF-300/US Activated ZIF-300

CR

Mass fraction / %

80

0 200

300

400 Temperature / oC

600

700 B

AN

6 5

M

4 3

ED

Deriv. weight / % oC -1

500

US

100

2 1

PT

0

200

CE

100

300

400 Temperature / oC

500

600

700

AC

Fig. 3. TGA-DTG curves: TGA (A) and DTG (B) of pure ZIF-300, pure Ultrason®, and ZIF300/US MMMs containing 10%, 20%, 30% and 40 % by mass of ZIF-300 Table 1.Characteristic temperatures of membrane materials acquired from TGA-DTG data Sample Designation Bare Ultrason® 10% ZIF300/US 20% ZIF300/US 30% ZIF300/US

Td5% /oC 175

Td10% /oC 450

Residual mass /% 1.9

1st DTG peak /oC 176

173

444

2.5

173

2nd DTG peak /oC 583 597 604

171

441

3.1

171 609

170

436

3.6

169 10

ACCEPTED MANUSCRIPT 40% ZIF300/US Pure ZIF-300

617 168 165

420 212

4.5 7.2

167 165

626

IP

T

As compared to unloaded Ultrason® membrane, thermal decomposition of MMMs taking place at elevated temperatures might be ascribed to enhanced filler-polymer interactions and improved thermal constancy of ZIF-300 nanocrystals. As maintained by 2nd DTG peaks, thermal stability of hybrid membranes was ameliorated by raising loadings of nanofiller and was found maximum for 40 wt % doping of ZIF-300 nanocrystals. The MMMs can be used free of danger in carbon capture applications because the highest temperature noticed in various combustion and gas separation processes falls in the range of 30-350 °C [22].

AC

CE

PT

ED

M

AN

US

CR

3.4 Gas Sorption Analysis N2 and CO2 adsorption isotherms obtained at various temperatures under relative pressure of ~00.1 MPa were used to determine some important physical macroscopic properties (such as density, nanofillers fractional volume etc.) and key microporous characteristics (such as total micropore volume, Brunauer-Emmett-Teller (SBET) and Langmuir (SLang) surface areas, N2 and CO2 uptakes, and CO2/N2 adsorption selectivity etc.) of fabricated membranes. Membranes microporosity and surface area were measured by using N2 adsorption isotherms obtained at 77 K. Loading capacities of N2 and CO2 gases in low pressure regime were assessed from N2 and CO2 adsorption isotherms at 298 K (Fig. 4). As outlined in Table 2, these properties are subsequently improved by increasing contents of incorporated ZIF-300 nanocrystals. The improved adsorption properties accurately corroborated the genuineness of uniform nanofiller dispersion and better adhesion at polymer-filler interface, thus resulting in fine quality composit membranes. The chabazite-type nanocrystals of ZIF-300 favorably adsorb quadrupolar CO2 gas molecules owing to their typical structural and chemical properties. This preferential adsorption significantly improves CO2 uptake for all MMMs in contrast to unfilled Ultrason® membrane. The maximum CO2 uptake of 0.7 mmol·g-1 (equivalent to 12 cm3·g-1) at 298 K was noted for the MMM containing 40 wt % ZIF-300 nanocrystals; the CO2 uptake can be further increased by lowering the operating temperature. Since the incorporated nanofillers have no specific chemical affinity for N2 gas, its uptake for all the prepared membranes followed almost a linear relationship with applied pressure up to 0.1 MPa. The size difference between two gas molecules resulted in controlled N2 adsorption as compared to CO2 on account of reduced pore size of ZIFfilled MMMs. As compared to N2, CO2 adsorption capacity of all the fabricated membranes was significantly high, particularly in low pressure regime. The CO2/N2 ideal adsorption selectivities of MMMs appreciably improved with increasing loadings of nanofiller; this increase is more sensitive at low partial pressure. The gas sorption study strongly recommends the use of these prepared MMMs in packed bed columns as efficient material for CO2 gas separation from post combustion flue gas where partial pressure of CO2 gas is comparatively low.

11

ACCEPTED MANUSCRIPT A

Activated ZIF-300 40% ZIF-300/US 30% ZIF-300/US 20% ZIF-300/US 10% ZIF-300/US Bare Ultrason

21

IP

T

14

7

CR

CO2 Uptake (cm3g-1)

28

0 0.2

0.4

0.6

US

0

0.8

1

P/P0

AN

ED

0.6

M

Activated ZIF-300 40% ZIF-300/US 30% ZIF-300/US 20% ZIF-300/US 10% ZIF-300/US Bare Ultrason

0.3

PT

N2 Uptake (cm3g-1)

1.2 0.9

B

0

0.2

0.4

0.6

0.8

1

P/P0

AC

CE

0

Fig. 4. CO2 and N2 adsorption isotherms for pure ZIF-300, pure Ultrason®, and ZIF-300/US MMMs containing 10%, 20%, 30% and 40 % by mass of ZIF-300 at 298 K

Table 2.Microporous properties of ZIF-300, Ultrason®, and ZIF-300/US MMMs Sample Designation

Density /g·cm-3

Фd /%

CO2 SBET SLang Vmicro 2 -1 2 -1 3 -1 uptake /m ·g /m ·g /m ·g /cm3·g-1

N2 uptake /cm3·g-1

CO2/N2 selectivity

12

ACCEPTED MANUSCRIPT 1.25

0

1.27

10

19.7

0.019

4.3

0.23

18

50

60

0.071

6.4

0.28

23

90

110

0.123

8.6

0.33

26

130

160

0.175

10.8

0.38

28

170

200

0.228

12.8

0.41

32

400

470

0.538

26.4

8.68 1.28 17.48 1.30

1.45

35.46 100

1.18

IP

1.31

T

26.40

22

CR

Pure Ultrason® 10% ZIF300/US 20% ZIF300/US 30% ZIF300/US 40% ZIF300/US Pure ZIF-300

AC

CE

PT

ED

M

AN

US

3.5 Gas Permeation Properties of MMMs 3.5.1 Dry and wet gas permeation Permeability, ideal selectivity, and coefficients of diffusion and solubility of fabricated membranes were determined via single gas (N2 and CO2) permeation experiments at 298 K and an upstream pressure of 0.2 MPa subjected to both dry and wet conditions. Inclusion of ZIF-300 nanocrystals resulted in substantial enhancements in CO2 permeability and CO2/N2 ideal selectivity as shown in Figs. 5 and 6 respectively. Permeation properties of CO2 molecules and their effective separation potentiality from N2 were not disturbed by the use of humidified gases in permeation experiments. The chabazite-type structural topology of ZIF-300 nanocrystals coupled with their noticeable chemical affinity for CO2 molecules over those of N2 greatly enhances CO2 permeability through the hybrid membrane. The optimum nanofillers composition rendering maximum throughput (measured in terms of CO2 permeability) and gas purity (quantified in terms of CO2/N2 permselectivity) was established to be 40 wt % ZIF-300 nanocrystals. CO2 permeability and CO2/N2 ideal selectivity of composite membrane associated with this specific loading was estimated to 27 Barrer and 26 respectively.

13

ACCEPTED MANUSCRIPT

25 20

CO2 CO2 N2 N2

15 10

IP

T

5 0

US

CR

Gas permeability /Barrer

30

AN

Membrane

PT

20

10 5 0

CE

15

AC

CO2/N2 Selectivity

25

Diffusion Solubility Ideal

ED

30

M

Fig. 5. Single gas CO2 and N2 permeability values of pure Ultrason® and ZIF-300/US MMMs with different ZIF loadings. (1 Barrer = 10-9 mol·m-2· s-1·Pa-1)

Membrane Fig. 6. CO2/N2 ideal selectivity values of pure Ultrason® and ZIF-300/US MMMs with different ZIF-300 loadings. 14

ACCEPTED MANUSCRIPT

T

3.5.2 Diffusivity and solubility of MMMs Coefficients of diffusivity (D) and solubility (S) computed for N2 and CO2 molecules penetrating through MMMs can help to understand gas permeation process through a membrane taking place by means of solution-diffusion mechanism. Figs. 7 and 8 respectively represent important facts related to gas diffusivity and solubility for the fabricated membranes. The increasing contents of ZIF-300 in MMMs lead to drop the diffusivity coefficients of both N2 and CO2 molecules as shown in Fig. 7. A significant drop in gas diffusivity coefficient took place due to pore size reduction of composite membranes caused by compact arrangement of ZIF-300 nanocrystals within the continuous polymer matrix.

IP

0.8

CR US

0.4

N2

0.2

AN

D×108 (cm2s−1)

0.6

CO2

PT

ED

M

0.0

Membrane

CE

Fig. 7. Pure gas coefficient of diffusion (D) for bare Ultrason® and ZIF-300/US MMMs having different ZIF-300 loadings.

AC

Normally the gas solubility coefficient mainly depends on structural arrangement and chemical formulation of nanofillers incorporated into polymer matrix. In addition, the presence of a large number of basic functional groups and interaction sites on ZIF-300 nanocrystals also improves the permselectivity of CO2 over N2. The increase in solubility coefficient of CO2 molecules was noted to be very high as compared to those of N2 molecules for every MMM specimen. The gas solubility coefficient significantly increased by increasing ZIF-300 contents and was maximized for the hybrid membrane containing 40 wt % ZIF-300 nanocrystals as depicted in Fig. 8. Owing to their specific chemical affinity for ZIF-300 nanocrystals, quadro-polar CO2 molecules, as compared to non-polar N2 molecules, are preferentially adsorbed onto ZIF crystals. This large solubility difference greatly enhances solubility of CO2 molecules as well as solubility-based CO2/N2 selectivity.

15

ACCEPTED MANUSCRIPT

CR

IP

T

Diffusion coefficients values for N2 and CO2 gases were estimated via time-lag method [Eq. (3)]. The enclosed term in Eq. (3) was corrected by making use of Langmuir parameters (K and y) which in turn were calculated from adsorption isotherms for N2 and CO2 gases acquired at 298 K. The values of diffusion coefficient obtained from Eq. (3) were used in Eq. (4) to calculate solubility coefficient values subjected to a pressure of 1 bar and temperature 298 K. An increase in ZIF-300 contents in polymer matrix led to reduce CO2/N2 selectivity (DCO2/DN2) based on diffusivity coefficient as described in Fig. 6. Gas selectivity (SCO2/SN2) based on solubility coefficient was increased by the addition of ZIF-300 nanocrystals as displayed in Fig. 6. The overall effect of gas permeation via solution-diffusion mechanism resulted in slight increase in ideal selectivity of CO2 over N2 [Fig. 6]. This improvement in permselectivity can be ascribed to the enlarged number of interaction sites and basic functional groups available on ZIF-300 nanocrystals.

CO2 N2

US

9

AN

6 3

M

S×102 (cm3 [STP] cm−3 atm)

12

CE

PT

ED

0

Membrane

AC

Fig. 8. Pure gas coefficient of solubility (S) for bare Ultrason® and ZIF-300/US MMMs having different ZIF-300 loadings.(1atm=101325Pa) The phenomenon of improved CO2/N2 permselectivity can also be elucidated in terms of reduced polymer inter-chain spacing. As compared to CO2 gas molecules, the adsorptive diffusion of N2 molecules was restricted owing to pore size constriction taking place within the polymer matrix. Pore size reduction takes place because ZIF-300 nanocrystals occupy and narrow down the polymer inter-chain void spaces. Both the restricted N2 diffusion and selective CO2 solubility taking place within the hybrid membrane result in improved CO2/N2 ideal permselectivity [Fig. 6]. The added nanofiller acted in such a way that the membrane structure provided precise openings for better CO2 permeability and CO2/N2 ideal selectivity. The comprehensive

16

ACCEPTED MANUSCRIPT investigation established on enhanced performance of ZIF-300/US MMMs firmly corroborated the development of a real mixed matrix membrane for CO2 capture from flue gas.

CR

IP

T

3.5.3 Comparison with Robeson upper bounds A chart plotted between CO2 permeability and CO2/N2 selectivity is used to compare the separation efficiency of prepared composite membranes with Robeson 1991- and 2008-upper bounds [23] as displayed in Fig. 9. In contrast to unfilled Ultrason® membrane, permselectivity values of composite membranes were significantly increased by the insertion of ZIF-300 nanocrystals into Ultrason® matrix. The permselectivity value of hybrid membrane containing 10 wt% ZIF-300 lies close to Robeson upper bound line while those of MMMs comprising 20wt%40 wt% ZIF-300 contents lie on this line, signifying their superiority over most of the current membranes [35-48], an important achievement of this work.

US M

AN

100

ED

10

PT

CO2/N2 Selectiviity

1000

10 100 CO2 Permeability

1000

AC

1

CE

1

Bare Ultrason 10% ZIF-300/US 20% ZIF-300/US 30% ZIF-300/US 40% ZIF-300/US Bare Matrimid ZIF-8/PSF HKUST-1/PSF MIL/PSF HKUST-1/Matrimid MOF-5/Matrimid HKUST-1/polyimide PIL Br-Matrimid ZIF-8/Matrimid S1C/PSF HKUST-1/S1C/PSF ZIF-8/S1C/PSF

Fig. 9. Comparison of CO2/N2 separation performance of ZIF-300/US MMMs with other MOF containing MMMs obtained from literature data. The Robeson upper bounds 1991 and 2008 for polymer separation performance are also shown. 3.6 Comparison of experimental data with mathematical models CO2 permeability through the fabricated composite membranes was estimated via two- and three-phase permeation models. As shown in Fig. 10, the two-phase Maxwell and Singh models established good match with experimental data at low filler loadings, whereas the three-phase Bruggeman and modified Felske models more closely predicted the permeation performance at 17

ACCEPTED MANUSCRIPT

T

IP CR

2 1

US

Relative permeability (Pr)

higher filler concentration. In contrast to three-phase models, relatively low experimental-vs.theoretical discrepancy associated with two-phase models indicates approximately ideal morphology of the prepared hybrid membranes. 5 Maxwell Bruggeman Singh 4 Modified Felske Experimental 3

0 0.1

0.2

0.3

0.4

AN

0

Filler Loading Fraction (Ф)

ED

M

Fig. 10. Comparison between experimentally determined CO2 permeability and those predicted by Maxwell, Bruggeman, Singh, and modified Felske models.

AC

CE

PT

Conclusions Polymer-based mixed-matrix membranes were developed by doping flexible Ultrason® polymer with ZIF-300 nanocrystals, having narrow particle size distribution, in varying loading levels. The prepared composite membranes were able to selectively and efficiently capture carbon dioxide from flue gas. The fabricated MMMs were found to be partially crystalline, thermally stable, microporous materials depicting consistent distribution of nanofiller and fine interfacial adhesion between polymer matrix and inorganic nanofiller as maintained by XRD, SEM, TGA and gas sorption experiments. As compared to unfilled Ultrason® membrane, CO2 permeability of composite membranes was enhanced by four times while the CO2/N2 ideal selectivity stayed almost unchanged. Moreover the presence of moisture contents in permeating gases did not upset separation performance of composite membranes. Furthermore, the carbon capturing effectiveness of fabricated MMMs was found close to the Robeson upper bound. The values of CO2 permeability and CO2/N2 permselectivity of composite membranes are high enough to fulfill industrial applications subjected to elevated temperatures. Acknowledgements The authors are thankful to KACST- Technology Innovation Center on Carbon Capture and Sequestration (CCS), King Fahd University of Petroleum and Minerals, Dhahran, Kingdom of Saudi Arabia (KSA) for providing support for this work. Nomenclature 18

ACCEPTED MANUSCRIPT

effective membrane area, cm2 diffusion coefficient, cm2·s-1 adsorption parameter determined from Langmuir adsorption isotherm membrane thickness, cm mass of the specimen gas permeability (Barrers, 1 Barrer = 10-9 mol·m-2· s-1·Pa-1) gas permeation rate (psi·s-1, 1 psi=6894.76Pa) in terms of time rate of pressure pressure difference across the membrane, psi(1 psi=6894.76Pa) universal gas constant (=6236.56 cm3·cmHg·mol-1·K-1)(1cmHg=13.3322Pa) solubility coefficient, cm3 (STP) ·cm-3·cmHg-1(1cmHg=13.3322Pa)

T V y α β γ δ θ λ ρ ψ Ф

absolute temperature, K cell downstream volume, cm3 parameter to be determined from Langmuir adsorption isotherm membrane gas selectivity matrix rigidification or chain immobilization factor ratio of interphase thickness to particle radius ratio of outer radius of rigidified interfacial matrix chain layer to radius of core particle x-rays diffraction angle, (o) permeability ratio density, g·cm-3 function of packing volume fraction of filler particles fractional volume of fillers, %

ED

M

AN

US

CR

IP

T

A D K L m P ΔP/dt Δp R S

AC

CE

PT

Subscripts C continuous phase D dispersed phase d5% 5 percent specimen mass loss d10% 10 percent specimen mass loss eff effective g glass transition i interphase i gas ‘i’ j gas ‘j’ m polymer matrix r relative

References [1] N. Mac Dowell, N. Florin, A. Buchard, J. Hallett, A. Galindo, G. Jackson, C. S. Adjiman, C. K. Williams, N. Shah, P. Fennell, An Overview of CO2 Capture Technologies, Energy Environ. Sci. 3 (2010) 1645-1669. [2] A. Hussain, M.B. Hägg, A feasibility study of CO2 capture from flue gas by a facilitated transport membrane, J. Membr. Sci. 359 (2010) 140-148. [3] R. Mahajan, W.J. Koros, Factors controlling successful formation of mixed-matrix gas separation materials, Ind. Eng. Chem. Res. 39 (2000) 2692-2696. 19

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

AN

US

CR

IP

T

[4] M. Rezakazemi,A.E. Amooghin, M.M. Montazer-Rahmati, A.F. Ismail, T. Matsuura, Stateof-the-art membrane based CO2 separation using mixed matrix membranes (MMMs): An overview on current status and future directions, Prog. Polym. Sci. 39 (2014) 817-861. [5] L.M. Robeson, Correlation of separation factor versus permeability for polymeric membranes, J. Membr. Sci. 62 (1991) 165-185. [6] S. Kim, L. Chen, J.K. Johnson, E. Marand, Polysulfone and functionalized carbon nanotube mixed matrix membranes for gas separation: Theory and experiment, J. Membr. Sci. 294 (2007) 147-158. [7] J. Ahn, W.-J. Chung, I. Pinnau, M.D. Guiver, Polysulfone/silica nanoparticle mixed-matrix membranes for gas separation, J. Membr. Sci. 314 (2008) 123-133. [8] B. Harold, T. Jeazet, C. Staudt, C. Janiak, Metal-organic frameworks in mixed-matrix membranes for gas separation, Dalton Trans. 41 (2012) 14003-14027. [9] B.D. Reid, A. Ruiz-Trevino, I.H. Musselman, K.J. Balkus, J.P. Ferraris, Gas permeability properties of polysulfone membranes containing the mesoporous molecular sieve MCM-41, Chem. Mater. 13 (2001) 2366-2373. [10] A. Car, C. Stropnik, K.-V. Peinemann, Hybrid membrane materials with different metalorganic frameworks (MOFs) for gas separation, Desalination 200 (2006) 424. [11] J. Liu, Y. Wang, A.I. Benin, P. Jakubczak, R.R. Willis, M.D. Le Van, CO2/H2O adsorption equilibrium and rates on metal-organic frameworks: HKUST-1 and Ni/DOBDC, Langmuir 26 (2010) 1430-14307. [12] B. Zornoza, S. Irusta, C. Tellez, J. Coronas, Mesoporous silica sphere-polysulfone mixed matrix membranes for gas separation, Langmuir 25 (2009) 5903-5909. [13] T.T.N. Nhung, H. Furukawa, F. Gandara, H.T. Nguyen, K.E. Cordova, O.M. Yaghi, Selective capture of carbon dioxide under humid conditions by hydrophobic chabazite-type zeolitic imidazolate frameworks, Angew. Chem. 126 (2014) 10821-10824. [14] A.C. Kizzie, A.G. Wong-Foy, A.J. Matzger, Effect of Humidity on the Performance of Microporous Coordination Polymers as Adsorbents for CO2 Capture, Langmuir 27 (2011) 63686373. [15] J. Liu, J. Tian, P.K. Thallapally, B.P. Mc Grail, Selective CO2 Capture from Flue Gas Using Metal Organic Frameworks- A Fixed Bed Study, J. Phys. Chem. C 116 (2012) 9575-9581. [16] Y. Wang, Y. Zhou, C. Liu, L. Zhou, Comparative studies of CO2 and CH4 sorption on activated carbon in presence of water, Colloids Surf. A 322 (2008) 14-18. [17] J.U. Wieneke, C. Staudt, Thermal stability of 6FDA-(co-)polyimides containing carboxylic acid groups. Polym. Degrad. Stab. 95 (2010) 684-693. [18] M. Sarfraz, M. Ba-Shammakh, Combined effect of CNTs with ZIF-302 into polysulfone to fabricate MMMs for enhanced CO2 separation from flue gases. Arab J Sci Eng. 41 (2016) 25732582. [19] D.R. Paul, D.R. Kemp, The diffusion time lag in polymer membranes containing adsorptive fillers, J. Polym Sci, Part C, Polym Symp. 41 (1973) 79-93. [20] H. Vinh-Thang, S. Kaliaguine, Predictive Models for Mixed-Matrix Membrane Performance: A Review, J. ACS Publications, 113 (2012) 4980-5028. [21] Y.C. Xiao, T.S. Chung, H.M. Guan, M.D. Guiver, Synthesis, cross-linking and carbonization of co-polyimides containing internal acetylene units for gas separation, J. Membr. Sci. 302 (2007) 254-264. [22] A.C.C. Chang, S.S.C. Chuang, M. Gray, Y. Soong, In-situ infrared study of CO2 adsorption on SBA-15 grafted with γ-(aminopropyl) triethoxysilane, Energy Fuels 17 (2003) 468-473. 20

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

AN

US

CR

IP

T

[23] L. M. Robeson, The upper bound revisited, J. Membr. Sc. 320 (2008) 390-400.

21

US

CR

IP

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

AN

Graphical abstract

22