Assessment of ZIF materials for CO₂ capture from high pressure natural gas streams

Highlights

• Metal organic frameworks ZIF-8, ZIF-14, and ZIF-71 synthesized for CO₂ capture from natural gas.

• High pressure isotherms measured and PSA simulations performed to assess performance.

• ZIF-14 unable to produce high purity methane.

• ZIF-8 and ZIF-71 produce high purity methane but unable to produce pure carbon dioxide.

• Metal organic frameworks studied have insufficient selectivity for CO₂ capture from natural gas.

Abstract

Although a considerable amount of the research is focussed in carbon capture specifically towards flue gas separations, another area that is of interest is CO₂/CH₄ separation for the natural gas industries. It is believed that 40% of the world’s reserves of natural gas are sour, and these gas reserves are typically left unexploited due to their high CO₂ content and the costs associated with separation and transport. One recent class of adsorbents, metal organic frameworks (MOFs) has been advocated as potential candidates for CO₂ removal from natural gas at high pressure. This can be attributed to their high CO₂ capacities, which could be exploited in high pressure separations. In this work we synthesised ZIFs -8, -14 and -71 and measured CO₂ and CH₄ isotherms over a range of temperatures and pressures. The CO₂ capacity for these materials at 303 K and 45 bar(a) was in the order of ZIF-8 (9.1 mol kg⁻¹) > ZIF-71 (8.1 mol kg⁻¹) > ZIF-14 (5.0 mol kg⁻¹). The CH₄ loading at 303 K and 100 bar(a) was in the order of ZIF-8 (6.8 mol kg⁻¹) > ZIF-14 (4.8 mol kg⁻¹) > ZIF-71 (4.4 mol kg⁻¹). The ideal selectivity of these materials for a 15%_mol CO₂, 85%_mol CH₄ feed mixture at 100 bar(a) and 303 K was found to be 5.6 for ZIF-8, 4.5 for ZIF-14 and 13 for ZIF-71. This isotherm data was then used to design and simulate a pressure swing adsorption process for CO₂/CH₄ separation.

Feed CO₂ concentrations between 15%_mol and 35%_mol were investigated at a condition of 100 bar(a) and 303 K. It was found that only ZIFs -8 and -71 could achieve the 98%_mol CH₄ product purity required. ZIF-8 and ZIF-71 were able to achieve CH₄ recoveries of 46%_mol and 48%_mol respectively. Furthermore, it was also found that ZIFs -8 and -71 behaved very similarly when compared on a volume of adsorbent basis. The CH₄ uptake of ZIF-14 was found to be abnormally high, which resulted in a very low CO₂/CH₄ selectivity. The loading of CH₄ was higher than CO₂ for both the 15%_mol and 25%_mol CO₂ feed cases, with only the 35%_mol CO₂ feed resulting in a higher CO₂ capacity, 4.6 mol kg⁻¹ of CO₂ in comparison to 4.1 mol kg⁻¹ of CH₄.

Although their CO₂ capacities at high pressures are high, there is little discrimination between the adsorption of small molecules. Consequently, their CH₄ loadings also increase substantially at those increased pressures. This results in a poor separation with the CO₂ product becoming diluted with the co-adsorbed CH₄ gas.

Introduction

Since their discovery in early 2006 [1], there have been many publications and innovations on zeolitic imidazolate frameworks (ZIFs). However, aside from the publication of Pérez-Pellitero et al. in 2010 [2], very little experimental data exists on the adsorption isotherms of CO₂ and CH₄. ZIFs can be classified as a variety of metal–organic framework (MOF), as they consist of metal centres linked by organic molecules and are named as such because their topologies mimic those of established zeolite structures. ZIFs traditionally utilise Zn or Co as the metal centre, however, a range of other metals, including, but not limited to Fe, Pr, Pb, Hg and Cd have been used [3]. In contrast to MOFs where a wide variety of organic molecules can be used as linkers, ZIFs are limited to imidazolate (Im) molecules, and their derivatives.

The ZIFs studied in this work include ZIF-8 (SOD topology), ZIF-14 (ANA topology) and ZIF-71 (RHO topology). A summary of these ZIFs are presented in Table 1. Phan [3] provides an overview of the range of ZIF materials currently available. ZIFs 8, 14 and 71 were chosen based on two main criteria, their pore apertures and their relative ease of synthesis and scale-up.

It should be noted that the aperture diameters do not necessarily restrict access to molecules of smaller diameter since larger molecules were found to adsorb on these materials. For example, Stepanov et al. [5], was able to adsorb benzene (kinetic diameter 5.85 Å) in ZIF-8, and in our current work, methane (kinetic diameter 3.8 Å), and nitrogen at 77 K were observed to adsorb in ZIF-14. The aperture values reported in Table 1 are determined based on a crystal structure refinement with the superposition of the van der Waals radii for the atoms. However, it is now reasonably understood for ZIF-8 that the organic linkers, or the structure, are not fixed in a given orientation [6], [7], [8], [9].

The Zn–Zn distance in ZIFs is approximately twice the distance of the Si–Si distance in a zeolite of the same topology [10]. Consequently, the pore volume of a ZIF is larger than that of the corresponding zeolite which gives rise to an increased pure gas adsorption capacity. There are other advantages to ZIFs over zeolites which include minimal water adsorption [11], [12] and according to simulation, ease of regeneration due to a low heat of adsorption [2]. The primary disadvantage of ZIFs is that their pore apertures cannot be modified as easily as their zeolite analogues. It is possible to modify the pore aperture and other adsorption characteristics of zeolites by extra-framework cation exchange or post synthesis modification.

The vast majority of the available literature on ZIFs consists mainly of the synthesis procedure, crystal structure solution and occasionally a surface area determination. Recently, a range of publications [2], [13] have reported a selection of adsorption isotherms for various gases, however, they are insufficient to conduct any process modelling. In addition, a recent publication by McEwen [14] compares the performance of ZIF-8, 13X and BPL activated carbon, however, isotherms were only measured at one temperature. Judging a material’s potential separation performance from a set of isotherms at a single temperature is generally unsuitable as the thermal effects during the process can be quite significant, as reported by Maring [15], and hence, with the currently available data, it is not possible to determine whether the selected ZIFs will be suitable for a given application.

The application investigated here is CH₄/CO₂ separation at high pressure, aimed towards the natural gas industries. Currently, when the CO₂ content of a field is high, the costs encountered in performing a bulk separation of CO₂ via the traditional amine absorption process is too high to justify the extraction, transportation and purification of the gas. It is estimated that up to 40% [16] of the known natural gas resources are not exploited due to this phenomenon. Hence, the materials studied here are evaluated on their ability to upgrade high CO₂ content gas fields with minimum valuable product (CH₄) loss.

In comparison to their zeolite counterparts where the CO₂ loading becomes saturated at low pressures, the continuously increasing loading of CO₂ on ZIF-8 at low pressures [14] suggests that this family of materials would show good performance in high pressure separations such as those that we are interested in this work. Hence, it was decided to investigate this family of materials further for their suitability and performance in such a situation. Furthermore, we found very little evidence of the experimental evaluation of ZIFs for the purposes of natural gas separation in the existing literature. The work conducted McEwen [14] on the CO₂–CH₄ separation on ZIF-8 for the purposes of natural gas, are insufficient to draw any conclusions on this matter as the range of experimental data lends itself to low pressure biogas separations. The aim of this work is to investigate natural gas separations at high pressure using ZIF materials. The work undertaken by Li [17] and Pérez-Pellitero et al. [2], also involving ZIF materials for CO₂–CH₄ separations do not cover the aspects that are encompassed in this in this work.

In this work, we synthesised ZIFs -8, -14 and -71, characterised them by XRD, TGA, SEM, nitrogen adsorption and measured adsorption isotherms for CO₂ and CH₄. The performance of the ZIFs was investigated over a range of CO₂ feed concentrations from 15%_mol to 35%_mol with the feed pressure and temperature fixed at 100 bar(a) and 303 K respectively. PSA performance over a range feed gas conditions was undertaken to evaluate the suitability of the three ZIFs for natural gas separation.