Zeolitic imidazolate framework membranes for gas separation: A review of synthesis methods and gas separation performance

Abstract

Zeolitic imidazolate framework (ZIF)-based membranes are emerging materials that are highly attractive for gas separation and purification. Highly reproducible defect-free ZIF membranes with well-defined pore structures can be formed in a number of ways. The performance of ZIF membranes depends largely on the preparation technique, application type, and operating conditions. Here, we review the methods of ZIF membrane preparation and discuss their performance in separating condensable and incondensable gases. In addition, recent developments in membrane synthesis and ways to improve the gas separation performance of ZIF membranes are discussed to promote industrial application and commercialization of ZIF membranes.

Introduction

Membrane technology is an attractive alternative for the purification of both incondensable inorganic gases (i.e. H₂, CO₂, N₂, O₂, SF₆, Ar, and CO) and condensable organic gases (i.e. CH₄, C₂H₄, C₂H₆, C₃H₆, C₃H₈, C₄H₁₀) to achieve a variety of separations, including those of isomers, chiral, and bio-molecules. In addition, membrane technology offers superior economics compared to conventional separation techniques, and it is estimated that the use of membranes can decrease energy costs by as much as 30% when compared with conventional, energy intensive technologies [1]. The cost-effectiveness of membrane technology is attributed to the ability to allow gasses to be treated in a single stage/phase as opposed to the multiple stages (pressurizing, depressurizing and purging) needed for conventional pressure swing adsorption systems or different phases (gas and liquid), which are common in commercialized cryogenic separation methods.

Despite their economic advantages, membrane technologies are not as widely used as other gas separation techniques due to manufacturing challenges. At present, only small and medium scale membrane units have been deployed due to the difficulty in achieving high permeability and the necessary level of throughput in large-scale industrial units. The materials that have been used to date consist primarily of dense polymeric membranes that exhibit high selectivity but low permeability; the limited performance of such polymeric membranes is well illustrated by Robeson's plot of selectivity versus permeability [2]. Intensive efforts to develop membranes that exceed Robeson's upper bound have been made by searching for appropriate structures and properties of membrane materials. As a result of these efforts, it has been determined that desirable membranes are those that exhibit high flux, high mechanical stability, and high selectivity [3]. Such membranes can be formed through metal–organic frameworks (MOFs), which consist of metal ions coordinated with organic molecules to form multi-dimensional structures with interconnected pores capable of sieving molecules and providing high flux, and which are suitable for use as large scale membranes in plants that require high throughput.

MOFs are organic–inorganic hybrid crystalline materials with well-defined pores and channels, have a porosity as high as 50% of the crystal volume, low density (0.2–1 g/cm³), and high surface area ranging from 1000 to 10,000 m²/g [4], [5]. MOFs are produced by reticular synthesis, which is the process of creating ordered structures (networks) with strong bonds between inorganic (metal ions) and organic units (organic linkers) [6]. The geometry of MOFs is designed by linking secondary building units (SBUs) with rigid shapes to obtain a predetermined network with high porosity, large pore openings, and functionalities [5], [7]. The reactivity of MOFs may be changed through post-synthetic modification (PSM) for catalytic purposes [8]. Multivariate MOFs (MTV-MOFs) that comprise different organic functionalities have also been synthesized [9]. The strong bonds that comprise MOFs (e.g. CC, CH, CO, and MO) are responsible for their high chemical and thermal stability, which extends from 250 to 500 °C [7], [10], [11], [12]. In addition, the high chemical stability of MOFs might permit their performance in water bearing processes that contain exhaust steam, such as the capture of CO₂ from the atmosphere or hot-flue gases produced at power plants, as well as for hydrogen and methane storage [4]. Gas separation by MOF films and membranes depends on the pore aperture size and the affinity of MOFs for the targeted gas [4]. Importantly, pore aperture size can be tuned by chemical functionalization of the organic ligand [1]. Catalysis is another field that can benefit from the high surface area, tunable pores, and high density of the active sites of MOFs. Specifically, as a result of these attributes, MOFs can be used for homogeneous catalyst support, short-lived catalyst stabilization, size selectivity, and catalyst encapsulation [4], [13].

MOFs are under investigation for their use in various biomedical applications such as drug storage and delivery, imaging, and sensing due to their high drug loading capacity, biodegradability, luminescence, and functionality [5], [14], [15]. MOFs can also be used for proton-conduction in fuel cells because their crystallinity can provide information about proton-conduction pathways and mechanisms, as well as new proton-conducting material design. These proton-conduction properties can be controlled by tuning the structure of MOFs [16].

Zeolitic imidazolate frameworks (ZIFs) are a subclass of MOFs that are topologically similar to zeolites. ZIFs are composed of tetrahedrally coordinated transition metal ions (e.g. Fe, Co, Cu, Zn) connected by organic imidazole linkers. ZIFs are emerging as materials that have the potential to improve the limited gas separation performance of polymeric membranes demarcated by Robeson's upper bound. The three-dimensional structures in ZIFs consist of tetrahedral transition metal cations (M) such as Co²⁺ and Zn²⁺, which are linked to imidazole-based (Im) ligands by coordination bonds. The bond angle of M-Im-M is similar to the SiOSi bond angle (145°) in zeolites. In these structures, the tetrahedral metal centers are coordinated by nitrogen atoms in the 1,3-positions of the imidazolate ligand (Im = C₃N₂H₃⁻) [17], and exhibit zeolite structures such as SOD, RHO, or LTA [18]. These porous, flexible, chemically, and thermally stable crystals (up to 400 °C) are promising materials for gas separation and storage, as well as for size- and shape-selective catalysis due to their pore size (less than 5 Å) [1], [10], [19]. In addition, their surface area is similar to that of very porous MOFs and is attributed to the full exposition of imidazole-based ligand edges and faces [10], [20]. ZIFs also show framework flexibility with respect to gas adsorption, and are stable under harsh conditions such as refluxing with organic solvents, water, and aqueous alkaline solutions [10].

The first reports of MOF [21], [22] and ZIF [23] membrane syntheses were quite recent, published in 2009. The synthesis of ZIF membranes is based on the classification of the synthesis methodology, which is briefly described and illustrated in Fig. 1. There are various methodologies for the synthesis of ZIF membranes; however, it is important to develop a synthesis strategy with low energy requirements that can be applied to the synthesis of all ZIF membranes with certain requirements such as controllable membrane thickness and reproducibility for effective gas separation. Such a general methodology may help to facilitate to the scaled-up processes necessary to achieve the ultimate goal of industrial application and commercialization of ZIF membranes. To promote this goal, we present a review of ZIF membrane synthesis methodologies and associated gas separation performances in laboratory settings. We also discuss innovative synthesis methodologies and how the challenges of ZIF membrane synthesis have been approached.