REVIEW pubs.acs.org/CR Carbon Dioxide Capture in Metal Organic Frameworks Kenji Sumida,† David L. Rogow,† Jarad A. Mason, Thomas M. McDonald, Eric D. Bloch, Zoey R. Herm, Tae-Hyun Bae, and Jeffrey R. Long* Department of Chemistry, University of California, Berkeley, California 94720-1460, United States 4.1.4. Non-CO2 Impurities in CO2/H2 Streams 4.1.5. Metal Organic Framework-Containing Membranes for Pre-combustion CO2 Capture 4.2. Metal Organic Frameworks as Adsorbents 4.2.1. Investigations Based on SingleComponent Isotherms 4.2.2. Computational Studies 5. Oxy-fuel Combustion 5.1. Metal Organic Frameworks for O2/N2 Separation 6. Metal Organic Framework-Containing Membranes 6.1. Continuous Films of Metal Organic Frameworks 6.2. Mixed-Matrix Membranes 7. Concluding Remarks and Outlook Author Information Biographies Acknowledgment List of Abbreviations References CONTENTS 1. Introduction 1.1. Carbon Dioxide Emission from Anthropogenic Sources 1.2. CO2 Capture at Stationary Point Sources 1.3. Options for CO2 Sequestration 1.4. Current CO2 Capture Materials 1.4.1. Aqueous Alkanolamine Absorbents 1.4.2. Solid Porous Adsorbent Materials 1.5. Metal Organic Frameworks 1.5.1. Synthesis and Structural Features 1.5.2. Physical Properties 2. CO2 Adsorption in Metal Organic Frameworks 2.1. Capacity for CO2 2.2. Enthalpy of Adsorption 2.3. Selectivity for CO2 2.3.1. Estimation from Single-Component Isotherms 2.3.2. Ideal Adsorbed Solution Theory (IAST) 2.3.3. Gas Mixtures and Breakthrough Experiments 2.4. In Situ Characterization of Adsorbed CO2 2.4.1. Structural Observations 2.4.2. Infrared Spectroscopy 2.5. Computational Modeling of CO2 Capture 3. Post-combustion Capture 3.1. Metal Organic Frameworks for CO2/N2 Separation 3.2. Enhancing CO2/N2 Selectivity via Surface Functionalization 3.2.1. Pores Functionalized by Nitrogen Bases 3.2.2. Other Strongly Polarizing Organic Functional Groups 3.2.3. Exposed Metal Cation Sites 3.3. Considerations for Application 3.3.1. Stability to Water Vapor 3.3.2. Other Minor Components of Flue Gas 4. Pre-combustion Capture 4.1. Considerations for Pre-combustion CO2 Capture 4.1.1. Advantages of Pre-combustion Capture 4.1.2. Hydrogen Purification 4.1.3. Metrics for Evaluating Adsorbents r 2011 American Chemical Society 724 725 726 727 727 728 729 731 731 732 733 733 733 739 741 742 742 742 743 744 745 746 756 756 756 757 757 760 761 762 763 767 770 771 771 772 772 774 1. INTRODUCTION The sharply rising level of atmospheric carbon dioxide resulting from anthropogenic emissions is one of the greatest environmental concerns facing our civilization today. These emissions, which stem predominantly from the combustion of coal, oil, and natural gas (ca. 80% of CO2 emissions worldwide),1 are projected to continue to increase in the future due to economic growth and industrial development, particularly in developing nations.2 Although the transition of the existing infrastructure from carbon-based sources to cleaner alternatives would be ideal in this regard, such a change requires considerable modifications to the current energy framework, and many of the proposed technologies are not yet sufficiently developed to facilitate largescale industrial implementation. Thus, carbon capture and sequestration (CCS) technologies that efficiently capture CO2 from existing emission sources will play a vital role until more significant modifications to the energy infrastructure can be realized. 746 746 746 749 750 752 752 754 754 755 755 755 755 Special Issue: 2012 Metal-Organic Frameworks Received: August 19, 2011 Published: December 28, 2011 724 dx.doi.org/10.1021/cr2003272 | Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW One scenario under which CO2 capture technologies could be rapidly implemented is at stationary point sources, such as coaland natural gas-fired power plants. In the United States, 41% of the total CO2 emissions can be attributed to electricity generation3 (ca. 60% worldwide),2 and hence the installation of effective capture systems to existing plant configurations could offer a large reduction in emissions. The captured CO2 would then be subjected to permanent sequestration, where the CO2 is injected into underground geological formations, such as depleted oil reservoirs or salt water aquifers. Here, similar technologies are already established in the context of processes such as enhanced oil recovery (EOR), and several trial CO2 sequestration sites are in construction.4 Note that the reuse of the captured CO2 as a reactant in chemical transformations presents an alternative sequestration pathway, although it would not be a viable long-term strategy owing to the tremendous scale of worldwide CO2 emissions (ca. 30 Gt per year)5 resulting in the market for any commodities prepared therefrom being rapidly saturated. A potentially more promising scenario that could utilize a considerable fraction of the captured CO2 would be its conversion into a fuel for transportation, provided efficient methods for carrying out the conversion via a renewable energy source can be developed.6,7 Nevertheless, regardless of the sequestration pathway, CCS systems must capture the CO2 from flue gas in an efficient, reversible fashion, and as will be discussed, the discovery of new materials exhibiting the right properties for performing CO2 capture is an area requiring urgent development. In this regard, the most significant challenge for CO2 capture at present is the large energy penalty associated with the capture process. With current technologies, approximately 70% of the cost of CCS is associated with the selective capture of CO2 from the power plant flue gas,8 a value that must certainly be reduced if such an approach for CO2 mitigation is to become viable. The high cost primarily arises from the large energy input required for regeneration of the capture material. Indeed, CO2 capture from a post-combustion flue gas using the most highly developed current technologies involving aqueous alkanolamine solutions carries an energy penalty of roughly 30% of the output of the power plant, most of which is associated with the liberation of the captured CO2 from the capture medium.8 Thus, minimization of the energy input for regeneration, through fine-tuning of the thermodynamics of the interaction between CO2 and the adsorbent, for example, is one of the most crucial considerations in improving the energy efficiency of CO2 capture. Metal organic frameworks are a new class of materials that could serve as an ideal platform for the development of nextgeneration CO2 capture materials owing to their large capacity for the adsorption of gases and their structural and chemical tunability.10,17,92 98 The ability to rationally select the framework components is expected to allow the affinity of the internal pore surface toward CO2 to be precisely controlled, facilitating materials properties that are optimized for the specific type of CO2 capture to be performed (post-combustion capture, precombustion capture, or oxy-fuel combustion) and potentially even for the specific power plant in which the capture system is to be installed. For this reason, significant effort has been made in recent years in improving the gas separation performance of metal organic frameworks, and some studies evaluating the prospects of deploying these materials in real-world CO2 capture systems have begun to emerge.9 16 Here, we review the progress that has been made in this area, with an emphasis on comparing the performance of metal organic frameworks to existing technologies, as well as highlighting the most crucial areas in which improvements in properties are urgently required. Note that, as a result of the rapid progress being made, we will limit the scope to CO2 capture from power plants. Other scenarios of CO2 capture, such as natural gas processing (CO2/CH4 separation), capture from transportation emissions, or direct air capture, are also highly important areas of research, and we direct the interested reader to a number of reviews that address a broader scope of gas separations for more detail.10,11,15,17 19 This review is intended to provide the reader with a comprehensive overview of the considerations associated with CO2 capture from power plants using metal organic frameworks and is arranged in the following manner. The remainder of section 1 provides an overview of the CO2 problem, a description of the current technologies utilized for CO2 capture from power plant flue gas streams, and an introduction to metal organic framework chemistry. Section 2 summarizes the various performance parameters to be evaluated in order to determine the performance of metal organic frameworks for CO2 capture applications and outlines a number of key characterization methods that are used for gaining a detailed correlation between the structural and chemical features of a metal organic framework and its adsorption properties. The potential utility of metal organic frameworks under the three main scenarios for CO2 capture from power plants, namely, post-combustion capture (section 3), pre-combustion capture (section 4), and oxy-fuel combustion (section 5) are then discussed. The emerging area of CO2 capture by membrane technologies, where the varying diffusion properties of the flue gas components are harnessed to effect gas separation, is presented in section 6. Finally, the outlook for utilizing metal organic frameworks for CO2 capture is discussed in section 7, noting areas in which further work is urgently required in order to realize next-generation metal organic frameworks that will be suitable for deployment within real-world systems. 1.1. Carbon Dioxide Emission from Anthropogenic Sources The escalating level of atmospheric CO2 has been welldocumented in recent times due to its implication in global warming, generating widespread environmental concerns toward the continued use of carbon-based fuels. The concentration of CO2 in the atmosphere at the present time is greater than at any other time in modern history, exceeding 390 ppm in 2011.20 To put this value into perspective, through analysis of ocean sediments for seawater pH and calcium, magnesium and carbonate mineralogy, and ice core data, it has been concluded that the atmospheric concentration has not approached such a level over the past 400 000 years.21,22 Indeed, over that period, prior to the extremely rapid increases observed over the past few decades, the atmospheric CO2 level has varied only gradually within the range 100 300 ppm. Thus, it is clear that global industrial development over the past century has created anthropogenic sources of CO2 leading to a rapid increase in the atmospheric concentration of CO2 to levels well above those to be expected from natural fluctuations. As shown in Figure 1, the combustion of fossil fuels represents over half of the global greenhouse gas emissions.2 Global CO2 emissions have increased by approximately 80% over the period 1970 2004 (from 21 to 38 Gt/year), and these emission levels are projected to increase further over the next several decades owing to rises in energy demands associated with a growing 725 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Figure 1. Global greenhouse gas emission sources in 2004 of which approximately 77% are represented by CO2 emissions.2 Figure 2. Projected electricity generation levels for coal and gas from 2007 to 2035.28 global population and economic and industrial development. The International Panel on Climate Change (IPCC) was formed in 1988 to assess the effects of greenhouse gas emissions on the global climate and predict possible future outcomes and to suggest remediation strategies.23 Based on the possible scenarios presented in the most recent IPCC climate report in 2007, the average global temperatures are expected to rise by between 1.8 and 6.4 °C by the end of the 21st century. Furthermore, current mitigation approaches are centered around the implementation of CO2 emissions pricing structures on the energy generation sector and a transition toward low- and non-carbon fuel sources. Among the strategies outlined, CO2 capture from fossil fuel-fired power plant emissions was deemed paramount to avoiding exacerbating climate change, since adaptation to low-carbon and alternative energy technologies, while urgent, will require many more years of research, development, and implementation. medium term until more environmentally sustainable energy sources can be deployed. The goal of the US Department of Energy/National Energy Technology Laboratory effort, which is being carried out as part of the Existing Plants, Emissions, and Capture Program, is to develop advanced CO2 capture and compression technologies for both existing and new coal-fired power plants that when combined can achieve capture of 90% of the CO2 produced at less than a 35% increase in the cost of electricity.2 Currently, the greatest challenge for the implementation of CO2 capture within power plants is the discovery of new materials that display suitable physical and chemical properties to be utilized within real-world systems that reduce the large energy requirements to perform the capture step. The composition of a typical post-combustion flue gas is shown in Table 1. Owing to the relatively low concentration of CO2 (15 16%) and the large quantities of N2 (73 77%) originating from the air in which the coal is combusted, a high selectivity toward CO2 is crucial, such that only pure CO2 is captured and subjected to sequestration. Current technologies involving aqueous amine absorbents capture the CO2 from the gas mixture with a high selectivity (see section 1.4.1 for further detail), but carry an energy penalty of approximately 30% of the power produced at the power plant.30 Here, the energy penalty originates primarily from the need to heat the large quantities of water in which the amine is dissolved, as well as the energy required to break the C N bond that is formed in the interaction between CO2 and the amine functionality. Thus, from the aspects of cost and environmental impacts of coal consumption, there is an urgent need to explore new materials that offer a lower energy penalty, ideally close to the thermodynamic minimum of approximately 10 20%.8 In the context of coal- or gas-fired power plants, there are three main scenarios under which new materials could serve to reduce the energy requirements of capture, as illustrated in Figure 3. In post-combustion capture (section 3), CO2 is removed from the flue gas that results after combustion of the fuel in air (see Table 1). This is predominantly a CO2/N2 gas separation owing to the high content of N2 in the air used for combustion and has been the most explored strategy to date since a post-combustion CO2 capture system could be readily retrofitted to existing power plants. Alternatively, pre-combustion capture (section 4) can be performed following gasification of the coal prior to combustion, 1.2. CO2 Capture at Stationary Point Sources Here, we review the main strategies that have been considered to date for CO2 capture from power plants. The CO2 emissions resulting from the combustion of coal for electricity generation represent 30 40% of the total anthropogenic CO2 contributions.24,25 In the United States, 43% of electricity production is derived from coal as of July 2011.26 In 2004, coal was responsible for 82.3% of the CO2 emissions originating from the electricity generation sector in the U.S.25 Coal-based power generation is expected to play an increasingly significant role in the future owing to the large reserves (>100 years) available for extraction,27 as well as the escalating global energy demands. For example, in China, which surpassed the United States as the largest emitter of greenhouse gases in 2006, the use of coal for electricity generation is projected to increase at a rate of 3.5% per year from 2.3 trillion kW 3 h in 2007 to 7.8 trillion kW 3 h in 2035.28,29 A similar trend is predicted at the global scale, where electricity generation from coal and gas is expected to increase from 11.8 trillion kW 3 h in 2007 to 21.9 trillion kW 3 h in 2035 (see Figure 2).28 Although the tremendous scale of coal and gas combustion for electricity generation is of significant concern, it also presents one of the most promising scenarios under which CO2 emissions could be dramatically reduced. Indeed, the installation of CO2 capture systems within coal- or gas-fired power plants that selectively remove the CO2 component of the flue gas would significantly reduce the global annual emissions, as well as lessen the environmental impact of an energy framework that will have an increased reliance on carbon-based fuels in the short to 726 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 1. Typical Postcombusion Flue Gas Composition for a Coal-Fired Power Plant31 molecule concentration (by volume) N2 73 77% CO2 H2O 15 16% 5 7% O2 3 4% SO2 800 ppm SO3 10 ppm NOx 500 ppm HCl 100 ppm CO 20 ppm hydrocarbons Hg 10 ppm 1 ppb producing a high-pressure flue gas flue gas containing H2 and CO2. This carries the advantage of being an easier separation than the CO2/N2 separation required for post-combustion CO2 capture. Once the CO2 is removed from the gas mixture, the H2 is then used for electricity generation, resulting in only H2O as the byproduct. Another possible method of reducing CO2 emissions would be to perform oxy-fuel combustion (section 5), in which pure O2 is utilized for the combustion of coal or natural gas. In this case, an O2/ N2 separation from air is performed, and the O2 is diluted with CO2 prior to combustion, leading to a flue gas that is a mixture of CO2 and H2O, which can be efficiently separated using existing technologies. Here, we note that each of the three processes requires a different gas separation, and there is a need for an entirely different set of materials properties for each separation due to the different physical properties of the gases, as listed in Table 2. This serves to highlight the importance of materials optimization, which will be essential in the development of next-generation separation materials. Figure 3. Basic schemes showing the types of CO2 capture relevant to the present review. The processes for post-combustion capture, pre-combustion capture, and oxy-fuel combustion, which are described in further detail in sections 3, 4, and 5, respectively. The main separation required for each type of process is indicated next to each of the headings in parentheses. and injected back into the well. A number of CCS demonstration projects are in the planning, implementation and operation stages around the world.37 The smallest example is a coal seam in Kaniow, Poland, where 760 t of CO2 was injected between 2004 and 2005. The largest project is the Erdos project in Inner Mongolia, China, where 3.6 million metric tons of CO2 per year is captured from a coal-to-liquid fuel plant and used for EOR. Storage in ocean water at depths of between 1000 and 3000 m, where CO2 is in the liquid form, is another option for sequestration of CO2. Potentially, up to 1 trillion tons of CO2 could be stored in deep ocean saline water as a result of injecting concentrated CO2 into deep water or creating CO2 pools on the ocean floor.38 Much of the CO2 could be stored on the ocean floor in the form of solid gas hydrates.39 It is well-known that enormous deposits of methane hydrates exist on the ocean floor. The ice-like solid complexes of water and CO2 formed under high pressure and cold temperatures may be an achievable form of storage that would be stable for long periods of time. 1.3. Options for CO2 Sequestration While the objective of this review is to discuss progress made in the area of the capture of CO2 from power plant emissions, we now briefly consider the aspect of CO2 storage (sequestration) of the enormous quantities of CO2 that would be obtained as a result of a successful capture framework. Although approximately 70% of the cost of CCS is derived from the capture step (a figure that is projected to decrease following future materials optimization), there are significant practical considerations associated with the other steps within a CCS system. For example, given that the quantities of CO2 that would be captured are too large for reuse by any chemical industry, the sustainable storage of CO2 is paramount of the successful development of the CCS. Currently, the most feasible scenario is one in which the CO2 removed from the power plant flue gas is, following compression, injected into an underground containment environment that facilitates CO2 storage without leakage and limited impact on the surrounding environment. The geological formations considered most suitable for long-term sequestration of CO2 are depleted oil and natural gas wells, shale, coal, and saline formations.34 Saline or brine-containing aquifers provide an environment where CO2 is able to react with mineral salts to form carbonates.35 In fact, injection of CO2 gas into oil and natural gas deposits can enhance the extraction of these fossil fuels by pressurization of the well in enhanced oil recovery (EOR) processes.36 In this case, the CO2 is usually recovered along with the oil or natural gas, then separated 1.4. Current CO2 Capture Materials The need for materials that can be used within CO2 capture systems for installation in coal- and gas-fired power plants has prompted the study of several classes of materials to date. The development of such materials requires the consideration of numerous performance parameters, which must be finely tuned depending on the type of CO2 capture and the specific configuration of the power plant. Optimization of these parameters should allow the energy penalty and cost of CO2 capture to be lowered, enabling widespread implementation under the various scenarios mentioned in section 1.2. The most crucial performance parameter for any CO2 capture material is its selectivity toward CO2. A high selectivity is 727 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 2. Physical Parameters of Gases Relevant to Carbon Dioxide Capture Processes10,32,33 kinetic diameter molecule (Å) dipole polarizability (10 25 cm 3) (10 esu 1 H2 2.89 N2 3.64 17.4 0 O2 3.46 15.8 0 CO 3.76 19.5 1.10 8.04 NO 3.49 17.0 H2O 2.65 14.5 0 cm 1) moment (10 27 esu 1 cm 1) 6.62 15.2 3.9 25.0 1.59 18.5 H2S 3.60 37.8 CO2 3.30 29.1 0 30.2 0 NO2 quadrupole moment 19 Scheme 1. Reaction of CO2 with Monoethanolamine (MEA) To Give a Carbamate Product (Upper), and the Corresponding Reaction with Triethanolamine (TEA) Resulting in a Bicarbonate Species (Lower) 9.78 43.0 20 30 wt %. Here, the reaction of 2 equiv of MEA with CO2 results in the formation of an anionic carbamate species and a corresponding ammonium cation. The total working capacity of a 30 wt % MEA solution is between 2.1 and 5.5 wt %, depending upon the specific configuration of the scrubbing process.42,43 In some cases, MEA is used in mixtures with secondary or tertiary alkanolamines, such as diethanolamine (DEA) and triethanolamine (TEA).30 In the case of TEA, the steric bulk about the nitrogen center of the tertiary amine results in the formation of a bicarbonate species rather than a carbamate species. Note that the stoichiometry of the reactions also has a significant impact on the maximum loading capacity for CO2, since the primary amines interact with CO2 in a 2:1 ratio, while the secondary and tertiary amines react in a ratio of 1:1. With regard to the enthalpy associated with the two reactions shown in Scheme 1, the bicarbonate species is comparatively less stable relative to the carbamate compound, resulting in a more readily reversible CO2 absorption reaction for the tertiary alkanolamine. Thus, the energy required to reverse the amine CO2 interaction generally decreases in the order of 1° > 2° > 3° amines, although the total regeneration energy within an actual capture system would depend on other factors, such as the concentration of the amine. Several other industrially relevant alkanolamines, including 2-amino-2-methyl-1-propanol (AMP) and N-methyldiethanolamine (MDEA), have also been investigated for use in CO2 absorbent solutions. Recently, some new amine-type solvents, piperazine44 and imidazolium-based ionic liquids,45 have gained much interest and have been shown to exhibit enhanced absorption properties in addition to a higher chemical and thermal stability compared with conventional amine solutions. For example, a 4:1 mixture of MEA amd MDEA used within a pilot plant exhibited a reduction in the overall energy requirement for CO2 capture compared with a system employing just MEA as the amine species in solution.46 Aqueous alkanolamine solutions have several significant limitations as adsorbents for large-scale CO2 capture. First, the solutions are relatively unstable toward heating, which limits the regeneration temperatures available for full regeneration of the capture material. Decomposition of the amine results in a decrease in absorbent performance over time, diminishing the lifetime of the solutions. The amine solutions are also corrosive toward the vessels in which they are contained, although this is usually prevented by the addition of corrosion inhibitors or by limiting the concentration of the alkanolamine species to below 40 wt %. Note that the latter is a key disadvantage for the use of alkanolamines, since a lower concentration results in a larger essential, such that the CO2 component of the flue gas is completely removed for subsequent sequestration. However, the affinity of the material toward CO2 is also a major consideration for optimizing the energy penalty of capture. Indeed, if the interaction is too strong, this leads to a high energy requirement for desorption of the captured CO2. On the other hand weak interactions, while lowering the regeneration cost, would afford low selectivities for CO2 over the other components of the flue gas. Furthermore, the material should exhibit a high stability under the conditions of capture and regeneration, such that it can be deployed for the lifetime of the power plant. Owing to the large quantities of CO2 that need to be removed from the flue gas, the materials should take up CO2 at a high density, such that the volume of the adsorbent bed can be minimized. In the following sections, we describe the main existing CO2 capture technologies in the context of these performance considerations, namely, aqueous alkanolamine solutions and porous solids, such as zeolites and activated carbons. As we shall see, none of the materials fulfill all of the criteria mentioned above, which highlights the urgent need for new materials to emerge that improve upon the characteristics of these materials. In this regard, metal organic frameworks represent an opportunity to create next-generation materials that are optimized for real-world applications in CO2 capture. However, the other types of materials also carry a number of advantages, and further optimization of these materials could also potentially allow development of suitable candidates that satisfy many of the performance criteria. 1.4.1. Aqueous Alkanolamine Absorbents. Aqueous alkanolamine solutions have been extensively studied to date for CO2 capture and are still considered the state-of-the-art despite being known for many decades.30 Here, the amine functionalities participate in a nucleophilic attack of the carbon atom of CO2 to form a C N bond, and depending on the amine, this results in the formation of a carbamate or bicarbonate species, as shown in Scheme 1.40 The affinity of these molecules for CO2 can be tuned to some extent by altering the substitution of the amine, although the strong orbital interactions usually require a high energy input for cleavage of this bond, and subsequent release of CO2. Indeed, the mechanism of CO2 adsorption in the case of alkanolamines falls into the chemisorptive regime, where the enthalpy of absorption lies the range of 50 to 100 kJ/mol at 298 K and low CO2 loadings.41 Monoethanolamine (MEA, Scheme 1, upper) is perhaps the most well-studied alkanolamine for CO2 capture applications and is usually dissolved in water at a concentration of approximately 728 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW zeolites coupled with their low cost and well-developed structural chemistry makes these materials an attractive target for use in CO2 capture applications. One feature of zeolites that may enhance their selectivity toward CO2 is the presence of charge-balancing metal cations within the pores. As will be discussed in the following sections, highly charged species on the surfaces of porous solids can afford high affinities for CO2 over other components of flue gas owing to the propensity for CO2 to be polarized to a higher extent compared with gases such as N2 or H2. For this reason, alkali and alkaline earth metal cation-exchanged chabazite-type zeolites were investigated as CO2 capture adsorbents.65 Indeed, it was found that the presence of exposed cations within the pores led to a higher affinity at low CO2 coverages, with the zero-coverage isosteric heat of adsorption falling in the range of 30 to 42 kJ/mol. The effect of such sites will be discussed in full detail when metal organic frameworks with exposed metal cation sites are considered in section 3. Activated carbons have also attracted much interest as CO2 adsorbents. These materials are amorphous porous forms of carbon that can be prepared by pyrolysis of various carboncontaining resins, fly ash, or biomass.9 The relatively uniform electric potential on the surfaces of activated carbons leads to a lower enthalpy of adsorption for CO2, and hence lower capacities for CO2 compared with zeolites at lower pressures. However, their significantly higher surface areas lead to greater adsorption capacities at high pressures, which has resulted in activated carbons being considered for a variety of high-pressure gas separation applications. In the context of CO2 capture, the high-pressure flue gas produced in pre-combustion CO2 capture has been a major target application for these materials. Indeed, one study has shown that the upper limit for the CO2 adsorption capacity within activated carbon materials is approximately 10 11 wt % under post-combustion CO2 capture conditions, while it reaches 60 70 wt % under pre-combustion CO2 capture conditions.66 A very recent study has also demonstrated that careful selection of the material precursors and the reaction conditions employed can lead to carbon-based adsorbents that have a volumetric CO2 adsorption capacity that is greater than some of the highest surface area metal organic frameworks at high pressure.67 One further advantage of activated carbons over zeolites is that their hydrophobic nature results in a reduced effect of the presence of water, and they consequently do not suffer from decomposition or decreased capacities under hydrated conditions.68 Moreover, consistent with the lower heat of adsorption for CO2, activated carbons require a lower temperature for regeneration compared with zeolites.68 Combining the high affinity of amines and the advantages of using a porous solid adsorbent holds tremendous promise for the development of new materials that exhibit properties that are appropriate for CO2 capture applications. New materials in which amines, alkanolamines, or alkylamines are grafted to the interior surfaces of activated carbons and zeolites are currently being surveyed experimentally.64,69 71 In one study, zeolite 13X was impregnanted with MEA at loadings ranging from 0.5 to 25 wt % by immersing the solid in methanol solutions of MEA.64 It was shown that an intermediate loading of 18.7 wt % was optimal for the selective adsorption of CO2 over N2, although the surface area decreased dramatically from 616 to 9 m2/g as a result of the amine molecules filling the pores of the zeolite. High-temperature (120 °C) adsorption studies found that the capacity for CO2 had increased from 4 to 14 mL/g, while at temperatures below 75 °C, the capacity was decreased. Figure 4. Heat capacities recorded for alkanolamine solutions47 and a representative metal organic framework, MOF-177.48 The heat capacity of the solid adsorbent is significantly lower, which is expected to result in a lower energy penalty for CO2 capture. volume of water that must be heated in order to regenerate the material. In fact, the high heat capacity of water (Cp = 4.18 J K 1 g 1) represents the main contribution to the regeneration energy costs.30 Indeed, as shown in Figure 4, the heat capacity of both 20 and 40 mol % MEA solutions are close to the heat capacity of pure water,47 and this is the main reason that solid porous adsorbents, which can have much lower heat capacities, are thought to present a promising strategy for reducing the regeneration energy penalty. 1.4.2. Solid Porous Adsorbent Materials. The lower heat capacity of solid porous adsorbents has led to their investigation as new materials for CO2 capture. In particular, zeolites,49 which are porous aluminosilicate materials that possess a high chemical and thermal stability, have been studied especially in the context of upgrading of natural gas and CO2 capture from post-combustion flue gas.50 54 For example, zeolite 13X, which has a relatively high surface area (SABET = 726 m2/g) and micropore volume (0.25 cm3/g), has been shown to display promising capacities for CO2 at room temperature (16.4 wt % at 0.8 bar and 298 K).55,56 The large variety of structures57,58 that have been reported to date presents an opportunity for the study of the effect of composition, or certain structural or chemical features, on the adsorption performance. For example, it has been observed that the Si/Al ratio of the material can have a significant impact on the CO2 adsorption within zeolite LTA, which in turn has crucial implications toward the adsorption selectivity and regeneration costs associated with the capture process.59 In comparison with post-combustion CO2 capture employing the alkanolamine solutions discussed above, small-scale pilot plants using zeolites have demonstrated more rapid adsorption of CO2 and lower energy penalty for the process.60 However, many of the zeolites studied to date become readily saturated with the water vapor present in the flue gas stream, and the CO2 adsorption capacity is consequently reduced over time.61,62 Furthermore, the large enthalpy of adsorption of CO2 leads to relatively high CO2 desorption temperatures (ca. 135 °C).63 This point highlights the importance of materials optimization for controlling the affinity of the pore surfaces toward CO2, and while improved synthetic procedures that provide a greater degree of control over the properties of the resulting zeolite have emerged, it still remains challenging to precisely tune the materials to the extent possible for metal organic frameworks. Nevertheless, the robust nature of 729 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Figure 5. Schematic diagrams of idealized temperature swing adsorption (TSA), pressure swing adsorption (PSA), and vacuum swing adsorption (VSA) processes for regenerating solid adsorbent in a fixed-bed column. by increasing the temperature (for TSA) or reducing the pressure (for PSA and VSA) of the bed. Due to the possibility of using low-grade heat from the power plant as a source of energy for regeneration, TSA is particularly promising for many CO2 capture processes.85,86 In a TSA cycle (Figure 5, left), the saturated adsorbent is heated from ambient pressure to the optimal desorption temperature of the material. As the temperature is increased, gas molecules desorb from the adsorbent surface, and the increased gas pressure drives the desorbed gases from the column. Once equilibrium is reached at the maximum desorption temperature and no more adsorbed gas elutes from the column, a purge is used to push off any desorbed gases that fill the void spaces of the bed until the purity of eluted gas falls below a desired level. Finally, the bed is cooled and prepared for the next adsorption cycle. For a PSA (Figure 5, center) or VSA (Figure 5, right) regeneration process, the column pressure is lowered after adsorption in order to desorb the captured gas. For PSA, the inlet gas is pressurized via compression, and flowed through the column until saturation. Once the inlet valve is closed, the column pressure decreases toward ambient pressure. The pressure drop desorbs significant quantities of adsorbates from the surface, which elute from the column. Similarly, VSA lowers the pressure of the column to subatmospheric pressure after adsorption at a higher pressure. The vacuum applied to the column removes the adsorbates from the pores. Since post-combustion flue gas is released near ambient pressure, compressing or applying a vacuum to such a large volume of gas is expected to be difficult, and therefore, TSA might represent the most viable process. For pre-combustion capture, the gas stream is inherently pressurized after the conversion reactions, and a PSA cycle is expected to be most appropriate. Indeed, selected metal organic frameworks have recently been evaluated in detail for use in post-combustion CO2 capture via TSA and a pre-combustion CO2 capture via PSA.48,87 The possibility of optimizing the parameters in each of these regeneration cycles (i.e., desorption temperature and pressure of inlet/outlet gas stream) and of combining multiple processes presents the option of tailoring the regeneration process to match the properties of a given adsorbent.88,89 Above all, regeneration strategies must be designed to minimize the total cost of capturing CO2, and as such, there will be a trade-off between maximizing the working capacity (the amount of CO2 that can be captured in a given adsorption cycle) and minimizing the energy required for regeneration.86,90,91 While detailed analysis of the energy and economic optimization issues in It was also demonstrated that the capacity decreased only slightly when the gas feed stream was saturated with moisture, compared with unmodified zeolite 13X. Carbamate species were identified in the infrared spectrum of the material at 120 °C, suggesting that the amines within the pores indeed do interact with CO2 in a chemisorptive fashion similar to that observed in the alkanolamine solutions. More recently, activated carbons and zeolite 13X were loaded with MEA and TEA by solution processing.71 Here, it was found that both materials exhibited selectivity for CO2 over CH4 (an important separation for natural gas sweetening) and N2 (postcombustion CO2 capture), although the CO2 capacity was decreased compared with the bare material owing to the lower surface area accessible to the gas molecules following installation of the amines. However, it was found that CO2 adsorption capacity increased with temperature from 298 to 348 K, which was attributed to an enhanced propensity for formation of carbonate species at the higher temperature. In addition to zeolites and carbon-based materials, other classes of porous materials are emerging as potential adsorbents in CO2 capture applications, including covalent-organic frameworks72 77 and amine-grafted silicas.78 80 In addition, microporous organic polymers81,82 have been shown to be of possible utility in gas separations relevant to CO2 capture.83,84 Here, the high mechanical processability of polymers (which is one of the significant advantages they hold over conventional crystalline porous materials) can be used to fabricate the material into the desired form, such as membranes. Furthermore, although the inherent selectivity of polymers is often limited, the fabrication of composite materials, where porous crystalline materials that are highly selective toward CO2 are embedded within the polymer phase, may afford new materials with enhanced properties. Nevertheless, significant work still needs to be performed to identify the most promising class of materials for CO2 capture, although the rich and highly diverse nature of the materials discovered recently holds tremendous promise for the discovery of next-generation materials that are optimized for this application. 1.4.2.1. Temperature and Pressure Swing Adsorption. In any CO2 capture process, the adsorbent must be regenerated after each adsorption cycle. Regeneration of a solid adsorbent is typically accomplished by temperature swing adsorption (TSA), pressure swing adsorption (PSA), vacuum swing adsorption (VSA), or some combination of these processes. In all cases, the solid adsorbent will likely be packed into a large fixed-bed column, and the adsorbate would be desorbed from the material 730 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW CO2 capture is beyond the scope of this review, such efforts will be crucial in directing the optimization of real-world CO2 capture processes. 1.5. Metal Organic Frameworks Metal organic frameworks are a new class of porous solids that have attracted much recent attention owing to their potential applications in a variety of areas, including gas storage, molecular separations, heterogeneous catalysis, and drug delivery.10,17,92 98 These structures consist of metal-based nodes (single ions or clusters) bridged by organic linking groups to form a one-, two-, or three-dimensional coordination network. From an applications point of view, their extraordinary surface areas,99 finely tunable pore surface properties,100 103 and potential scalability to industrial scale104 have made these materials an attractive target for further study. Although a number of excellent reviews already exist on the synthetic and structural aspects of metal organic frameworks,92,93,105,106 we briefly describe a number of key aspects of these materials to aid the reader in grasping the promises and challenges of utilizing metal organic frameworks for CO2 capture. 1.5.1. Synthesis and Structural Features. The synthesis of metal organic frameworks is conventionally achieved by employing a so-called modular synthesis, wherein metal ions and organic ligands are combined to afford a crystalline, porous network. A large variety of synthetic methods have been developed in recent years for the preparation of these materials, and the conditions that lead to the formation of the desired phases are widely variable. Indeed, the reported synthetic procedures encompass a wide range of temperatures, solvent compositions, reagent ratios, reagent concentrations, and reaction times, and the fine-tuning of all of these parameters is crucial in optimizing the synthesis of the materials. Microwave heating,107 sonicationassisted synthesis,108,109 or mechanochemical procedures110,111 have also been employed to effect formation of the framework. More recently, the first example of electrosynthetic deposition of metal organic frameworks has also been demonstrated.112 In most of the resulting materials, the solvent used during synthesis occupies the void space within the pores but can be removed by application of vacuum or heat, resulting in a large pore volume and large surface area accessible to guest molecules. In the context of CO2 capture, these surfaces can then be utilized to perform the gas separation. One of the most well-studied metal organic frameworks to date is the Zn4O(BDC)3 (MOF-5) compound depicted in Figure 6,113 115 which consists of tetrahedral [Zn4O]6+ clusters bridged by ditopic BDC2 ligands to form a cubic, threedimensional network. Importantly, functionalized derivatives of the MOF-5 structure type can be prepared using other substituted linear dicarboxylate linkers, allowing the linker length or functional groups present on the aromatic backbone to be readily modified, while preserving the overall connectivity of the framework.100,116 This family of materials, referred to as isoreticular metal organic frameworks (IRMOFs), feature considerably different pore sizes and pore functionalities, suggesting that the properties of the materials can be finely tuned by employing the appropriate linker type. More recently, this concept has been extended to other families of materials, such as the Zr6O4(OH)4(BDC)6 (UiO-66),117,118 Al(OH)(BDC) (MIL-53),119,120 and Cu2(BPTC) (NOTT-100)121 structure types. In each case, the length and functionality of the ligands can be changed, while forming materials of the same network Figure 6. A portion of the crystal structure of Zn4O(BDC)3 (MOF-5). Blue tetrahedra represent ZnO4 units, while gray and red spheres represent C and O atoms, respectively; H atoms are omitted for clarity. connectivities. In the case of the MIL-53 structure type, there is also a dynamic (flexible) behavior, which is dependent on the constituents of the material. Note that the flexibility of metal organic frameworks is an area that has also begun to be explored in the context of providing high selectivity,122,123 although, as we discuss in section 3, the evaluation of their performance is more complex compared with rigid structures. Nevertheless, in the context of CO2 capture, the ability to readily modify the surface chemistry of metal organic frameworks is of particular interest for installing the desired chemical features (such as amines or polarizing groups) for enhancing the performance of the material. An important route to installing the desired functionalities on the organic bridging unit is the postsynthetic modification of the surface functional groups following the initial formation of the crystalline structure.102,103 One advantage of such an approach is that functional groups that might interfere with the formation of the framework owing to their propensity to bind metal ions (such as amines, alcohols, and aldehydes) can be installed with wellknown organic transformations once the framework scaffold has been formed, eliminating the need to develop precise reaction conditions to form the material directly. Such a procedure has been demonstrated on compounds such as IRMOF-3,124 DMOF-1-NH2,125,126 UiO-66-Br,118 and MIL-101(Cr),127 and the scope of reactions available is growing rapidly. An alternative strategy of preparing metal organic frameworks with functionalized surfaces is the generation of materials possessing exposed metal cation sites in the pores.128 132 As will be discussed in section 3.2.3, such sites can facilitate highly selective interactions with certain gas molecules in separation applications and can also be of benefit for high-density gas storage applications, since they permit close approach of guest molecules to the pore surface. One of the most studied materials featuring such binding sites is Cu3(BTC)2 (HKUST-1), which exhibits a cubic, twisted boracite topology constructed from dinuclear Cu2+ paddlewheel units and triangular 1,3,5-benzenetricarboxylate linkers (see Figure 7).133 The as-synthesized form of the framework contains bound solvent molecules on the axial coordination sites of each Cu2+ metal center, which can be 731 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Figure 8. A portion of the crystal structure of M2(dobdc) (M-MOF-74 or CPO-27-M). Black, gray, and red spheres represent M, C, and O atoms, respectively; H atoms are omitted for clarity. Figure 7. A portion of the crystal structure of Cu3(BTC)2 (HKUST-1).133 Green, gray, and red spheres represent Cu, C, and O atoms, respectively; H atoms are omitted for clarity. solution pH, metal-to-ligand ratio, metal counteranion, reaction temperature, and reaction time, can have a considerable impact on the products that are obtained. Note that even in cases where just a single metal and ligand are combined, a large number of network connectivities may be possible, many of which are nonporous structures that are not of interest for gas storage and gas separation applications. These undesired phases can often coprecipitate with the porous phase of interest, giving rise to a mixture that is not readily separated owing to the insolubility of the components. Thus, the discovery of optimized conditions that afford the desired product can involve a large number of trial reactions in which the reaction parameters are systematically varied. Consequently, numerous studies involving the use of high-throughput technologies have appeared in recent years.140,151 154 1.5.2. Physical Properties. The chemical and thermal stability of metal organic frameworks is generally lower than that of zeolites and other porous inorganic solids due to relatively weak coordination bonds that connect the metal and ligand components. Many are especially air- and moisture-sensitive following evacuation of the pores, leading to the need for careful handling under an inert atmosphere if the best performance characteristics are to be obtained. In the case of MOF-5, even slight exposure of the activated form of the material to the air results in rapid degradation of the crystallinity of the material and a concomitant loss of its surface area owing to the hydrolysis of the Zn O bonds.115,155 For applications such as post-combustion CO2 capture, where significant quantities of water are present in the incoming gas stream, an increased chemical stability will be crucial if metal organic frameworks are to be employed as separation materials. Several approaches for improving the stability of metal organic frameworks (mainly through the use of components expected to yield stronger metal ligand bonding) will be addressed in section 3.3.1. The mechanical properties of metal organic frameworks is an aspect that has not been thoroughly explored compared with organic polymers, though recent progress has been made in correlating the chemical structure to the observed characteristics.156 For CO2 capture applications, the materials should be sufficiently mechanically stable to allow a dense packing of the adsorbent bed without loss of the network structure. Even slight perturbations to the structural or chemical features under a high mechanical subsequently removed in vacuo at elevated temperatures to create open binding sites for guest molecules. Indeed, these sites act as charge-dense point charges, which provide an opportunity for discrimination of certain components of gas mixtures based on their polarizability and dipole or quadrupole moment. Interestingly, the HKUST-1 structure type can be prepared using a variety of metal ions (M = Cr, Fe, Zn, Mo).134 137 In fact new members of several structure types, such as M2(dobdc), M-BTT,138 140 MIL-53,119,141 147 and MIL-88,148 have been prepared following their initial reports, highlighting that the metal is an additional variable worthy of consideration when tuning the properties of the material. In cases where the metal ions possess exposed coordination sites, the adsorptive properties can be dramatically tuned based on the charge density present at the metal coordination site. In the context of CO2 capture, the M2(dobdc) (M = Mg, Mn, Fe, Co, Ni, Zn) series of metal organic frameworks displayed in Figure 8 have recently been intensively studied.100,130,149,150 These compounds exhibit a honeycomb-type network topology featuring hexagonal one-dimensional channels with a high density of exposed M2+ adsorption sites. As discussed in section 3, the specific metal that is employed has a dramatic effect on the adsorption capacity, binding strength, and material stability, highlighting the importance of materials optimization for specific applications. In the case of the Fe2(dobdc), the ability for the exposed Fe2+ cations to engage in redox chemistry has allowed selective interactions with O2 over N2, which is of importance for oxy-fuel combustion (see section 4). Nevertheless, the systematic study of such a large family of isostructural compounds allows the effect of changing the metal ion to be identified, and more studies of this type would allow a greater understanding of the role of the metal ion in facilitating the adsorption properties of metal organic frameworks. Despite the apparent simplicity of the modular synthesis of metal organic frameworks, one of the greatest challenges in preparing new materials lies in optimizing the reaction conditions that lead to the desired metal organic framework in high yield and crystallinity. Slight changes in the reaction parameters employed, such as the reactant concentration, the presence of a cosolvent, 732 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW have recently emerged as invaluable tools for probing the adsorption phenomena. In this section, we present the most significant results that have been reported in this regard. pressure could have a considerable effect on the overall performance of the capture system. In terms of the magnitude of the structural changes that have been observed to date, it has been demonstrated for the metal organic framework Cu3(BTC)2 that the application of large pressures (on the order of several GPa) can lower the total lattice volume of the material by as much as 10%.157 It is not yet apparent how this compression would affect the adsorption properties of the material, but such effects would certainly be worthy of investigation when evaluating metal organic frameworks within a real-world CO2 capture system. An additional bulk property that should be considered when assessing metal organic frameworks for installation within a packed bed is its thermal conductivity. This parameter is crucial in determining the heating efficiency of the adsorbent bed and the duration of the regeneration cycle of a temperature swing adsorption-based capture process. Some analyses of this type have been performed in recent years, although owing to the limited scope of the studies, it is not yet established how much variation might exist in the thermal conductivity between compounds of different compositions and structure types. In a study of MOF-5,158 the thermal conductivity was shown to decrease rapidly at the lowest temperatures (20 100 K), but remain relatively constant above 100 K. The thermal conductivity value of 0.32 W 3 m 3 K is higher than that measured for zeolite NaX,159 although the significant effect of particle size (larger particles exhibit higher thermal conductivities) presents a difficulty in comparing two materials. Nevertheless, the values observed for MOF-5 suggest that the thermal conductivity would not be a barrier for the implementation of metal organic frameworks within an adsorbent bed. One particular advantage of the use of a solid adsorbent such as a metal organic framework would be the greatly reduced heat capacity compared with the aqueous alkanolamine solutions discussed in section 1.4.1. This is a particularly important parameter in temperature swing adsorption-based processes since it reflects the quantity of energy required for heating of the sorbent material to the desorption (regeneration) temperature. Indeed, as shown in Figure 4, the heat capacity of Zn4O(BTB)2 (MOF-177) over a temperature range of 25 200 °C was determined to be between 0.5 and 1.5 J 3 g 1 3 K 1, which lies considerably below the corresponding values recorded for 20 and 40 wt % aqueous MEA solutions.47 Thus, if robust metal organic frameworks exhibiting suitable gas separation performance can be prepared, there is tremendous promise in realizing CO2 capture processes with greatly reduced energy requirements. 2.1. Capacity for CO2 The adsorptive capacity is a critical parameter for consideration when evaluating metal organic frameworks for CO2 capture. The gravimetric CO2 uptake, which refers to the quantity of CO2 adsorbed within a unit mass of the material, dictates the mass of the metal organic framework required to form the adsorbent bed. Meanwhile, the volumetric capacity refers to how densely the CO2 can be stored within the material and is an equally crucial parameter, since it has a significant influence on the volume of the adsorbent bed. Both parameters also have an important role in determining the heating efficiency of the metal organic framework, which directly impacts the energy penalty required for material regeneration and desorption of the captured CO2. The high internal surface areas of metal organic frameworks provide an opportunity for large CO2 adsorption capacities to be achieved, owing to the efficient packing and close approach of the guest molecules on the pore surface. For example, at 35 bar, the volumetric CO2 adsorption capacity for MOF-177 reaches a storage density of 320 cm3(STP)/cm3, which is approximately 9 times higher than the quantity stored at this pressure in a container without the metal organic framework and is higher than conventional materials used for such an application, namely, zeolite 13X and MAXSORB.160 Note that, as mentioned earlier, the volumetric adsorption capacity of activated carbon materials at high pressures has been demonstrated to be competitive with metal organic frameworks,67 although we do not consider those materials in any further detail in the following discussion. The high-pressure adsorption capacities for selected metal organic frameworks are tabulated in Table 3. Indeed, the greatest capacities at high pressures are observed for materials exhibiting large surface areas, although excellent adsorption properties have also been demonstrated in a number of materials with modest surface areas that have a significant density of high-affinity adsorption sites, such as exposed metal cations. Owing to the particular relevance to pre-combustion CO2 capture, which is performed at high pressures, these data will be discussed in full detail in section 4. The lower-pressure (<1.2 bar) adsorption capacities for metal organic frameworks collected at ambient temperatures (293 313 K) and 273 K are presented in Tables 4 and 5, respectively. At these pressures and temperatures, the adsorptive properties are predominantly dictated by the chemical features of the pore surface, and most of the high-capacity materials are those bearing highly functionalized surfaces. For post-combustion CO2 capture, the pressure of the flue gas (∼1 bar) and the low partial pressure of CO2 thereof (PCO2 ≈ 0.15 bar) leads to the lowerpressure region of the CO2 adsorption isotherm being the area of interest (see section 3). Thus, maximizing the adsorption capacity specifically around 0.15 bar would be expected to lead to new materials with enhanced performance for post-combustion CO2 capture. 2. CO2 ADSORPTION IN METAL ORGANIC FRAMEWORKS In this section, we address a number of aspects related to the adsorption of CO2 within metal organic frameworks that are important considerations when evaluating new materials for CO2 capture applications, such as the adsorption capacity and enthalpy of adsorption. Single-component gas adsorption isotherm data can further be used to estimate the adsorption selectivity for CO2 over other gases, which is a crucial parameter that determines the purity of the captured CO2. Detailed knowledge of the binding environment of CO2 within the pores of the framework can give vital information regarding the structural and chemical features contributing to the observed material performance, and in situ vibrational spectroscopy and crystallographic methods 2.2. Enthalpy of Adsorption The enthalpy of adsorption of CO2 is a critical parameter that has a significant influence over the performance of a given material for CO2 capture applications. The magnitude of the enthalpy of adsorption dictates the affinity of the pore surface 733 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 3. High-Pressure CO2 Adsorption Capacities in Selected Metal Organic Frameworks between 273 and 313 K surface area (m2/g) chemical formulaa common name BET Langmuir capacityb (wt %) pressure (bar) temp (K) ref Zn4O(BTE)4/3(BPDC) MOF-210 6240 10400 74.2 50 298 Zn4O(BBC)2(H2O)3 MOF-200 4530 10400 73.9 50 298 99 Cu3(TCEPEB) NU-100 6143 69.8d 40 298 161 69.0d 28 300 162 68.9c 36 278 131 2060 39.8 40 313 87 2900 65.0d 10 273 163 4400 2900 58.5 58.0d 40 10 313 273 87 163 3008 45.1d 40 298 164 6170 62.6 50 298 99 62.1c 47 298 165 99 Zn4O(FMA)3 Mg2(dobdc) 1120 Mg-MOF-74, CPO-27-Mg 1542 1800 Zn3.16Co0.84O(BDC)3 Co21-MOF-5 Be12(OH)12(BTB)4 Zn4O(BDC)3 Be-BTB MOF-5, IRMOF-1 Zn4O(BTB)4/3(NDC) MOF-205 Ni5O2(BTB)2 DUT-9 Zn4O(BTB)2 MOF-177 [Cu3(H2O)]3(ptei) Cr3O(H2O)2F(BDC)3 PCN-68 MIL-101(Cr) 1618 4030 4460 99 4500 5340 60.8 50 298 4690 5400 60.6 40 313 87 4898 6210 56.8d 30 298 108 5109 4230 6033 57.2c 56.9d 35 50 298 304 166 167 3360 4792 50.2d 30 298 168 54.2c 22 278 131 Ni2(dobdc) Ni-MOF-74, CPO-27-Ni 1218 [Cu(H2O)]3(ntei) PCN-66 4000 4600 53.6c 35 298 166 Zn4O(BDC)(BTB)4/3 UMCM-1 4100 6500 52.7c 24 298 169 [Cu(H2O)]3(btei) PCN-61 3000 3500 50.8c 35 298 166 Cu4(TDCPTM) NOTT-140 2620 46.2d 20 293 170 Tb16(TATB)16(DMA)24 Cr3O(H2O)3F(BTC)2 MIL-100(Cr) 1783 1900 44.2d 44.2d 43 50 298 304 171 167 Cu3(BTC)2 HKUST-1 1270 42.8 300 313 172 2211 40.1 40 303 173 3855 35.9 15 298 174, 175 H3[(Cu4Cl)3(BTTri)8] Cu-BTTri 1571 1750 2050 42.8 40 313 87 Co(BDP) Co-BDP 2030 2780 41.3 40 313 87 37.6 15 298 176 35.0d 30.6d 30 25 298 304 177 178 Zn2(BDC)2(DABCO) Zn(MeIm)2 Al(OH)(BDC) nZif-8, n = nano MIL-53(Al) Al(OH)(BDC-NH2) amino-MIL-53(Al) Zn2(BPnDC)2(bpy) SNU-9 Al(OH)(ndc) DUT-4 Cr3O(H2O)2F(BDC)3 MIL-101(Cr) Zn(F-pymo)2 β-Zn(F-pymo) Mn(tpom)(SH)2 MSF-2 Zn6O4(OH)4(BDC)6 Zn(Gly-Ala)2 1264 1308 30.0d 30 303 147 1030 29.9c 30 298 179 1996 26.4 10 303 180 283 181 26.0d 5.3 26.0d 28 273 182 25.0d 20.3 293 183 UiO-66 24.3d 19.0d 18 15 303 273 184 185 Al12O(OH)18(H2O)3(Al2(OH)4)(BTC)6 MIL-96(Al) 18.6d 20 303 186 Cr3O(H2O)2F(NTC)1.5 MIL-102 13.0d 30 304 187 622 816 a See List of Abbreviations. b Adsorption capacities are calculated from absolute adsorption isotherms unless otherwise noted and estimated from adsorption isotherms in cases where the values were not specifically reported. c Reported capacity was calculated from the excess adsorption isotherm. d It was not clear from the reference whether the reported isotherms were in absolute or excess adsorption. CO2 too strongly would increase the regeneration cost owing to the large quantity of energy required in order to break the framework CO2 interactions. Meanwhile, if the enthalpy of adsorption is too low, although the material would be more readily regenerated, the purity of the captured CO2 would be lowered owing to the decreased adsorption selectivity, and toward CO2, which in turn plays a crucial role in determining the adsorptive selectivity and the energy required to release the CO2 molecules during regeneration. Precise control of the binding strength of CO2 is essential if metal organic frameworks are to be optimized such that they can lower the energy requirements of the capture process. Specifically, the use of a material that binds 734 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 4. Lower-Pressure CO2 Adsorption Capacities for Metal Organic Frameworks at 293 319 K surface area (m2/g) chemical formulaa Mg2(dobdc) common name Mg-MOF-74, CPO-27-Mg Cu3(BTC)2(H2O)1.5 HKUST-1, (4 wt % H2O) Co2(dobdc) Co-MOF-74 BET 1174 Langmuir 1733 capacityb (wt %) pressure (bar) temp (K) ref 27.5 1 298 27.2 1 298 188 189 298 48 1800 2060 26.7 1 26 1.1 298 190 1495 1905 26 1 296 150 27 1 298 191 957 1388 24.9 1 298 189 CPO-27-Co 1080 Ni2(dobdc) Ni-MOF-74 936 23.4 1 296 150 1356 23.9 1 298 189 22.7 1 298 192 1312 22.6 1 303 193 20.4 1 296 150 19.8 1 296 189 19.6 1 296 150 17.6 1.1 298 160 CPO-27-Ni 639 1083 1070 Zn2(dobdc) Zn-MOF-74 CPO-27-Zn 816 Cu3(BTC)2 HKUST-1 1400 1492 HKUST-1, (8 wt % H2O) H3[(Cu4Cl)3(BTTri)8(mmen)12] mmen-Cu-BTTri Co2(adenine)2(CO2CH3)2 H3[(Cu4Cl)3(BTTri)8] Zn2(ox)(atz)2 CuTATB-60, PCN-6 1 295 189 1 298 160 1482 15 10.6 1 0.8 295 298 195 192 6.2 1 313 196 6.2 17.4 1 1 295 298 195 191 197 3811 15.9 1 298 870 15.4 1 298 198 bio-MOF-11 1040 15.2 1 298 199 Cu-BTTri 1770 782 1900 14.3 14.3 1 1.2 298 293 200 201 14 1 298 146 2450 13.7 1 298 202 13.5 1 298 140 3065 13.4 1 298 197 203 USO-2-Ni-A 1530 SNU-50 2300 Fe3[(Fe4Cl)3(BTT)8(MeOH)4]2 Fe-BTT 2010 Cu3(TATB)2 Cu-TATB-30 2665 Cu(bpy)2(BF4)2 ELM-11 Cu2(bptc)(H2O)2(DMF)3 MOF-505 Al(OH)(2-amino-BDC) Al(OH)(bpydc) 3 0.97Cu(BF4)2 NH2-MIL-53(Al), USO-1-Al-A NOTT-140a Cu2(TCM) SNU-21S UMCM-150(N2) Pd(μ-F-pymo-N1,N3)2 MIL-53(Al), USO-1-Al 12.7 1 298 12.6 1.1 298 160 960 705 12 11.8 1 1 298 298 146 204 2620 Cu3(BPT)2 USO-2-Ni Ni-STA-12 Ni3(pzdc)2(7H-ade)2(H2O)4 11.7 1 293 170 11.7 1 298 205 11.1 1 298 206 10.8 1 298 189 600 10.7 0.86 293 207 1300 10.6 1 298 146 9.2 10.2 1 1 303 298 208 189 UMCM-150 Ni2(pbmp) 1170 905 1235 Ni2(BDC)2(DABCO) 4436 1547 886 Cu3(BPT(N2))2 Al(OH)(BDC) 194 191 15.2 Ni2(2-amino-BDC)2(DABCO) Co(tImb) 3 DMF 3 H2O 293 298 18.3 Cu2(bdcppi)(DMF)2 Cu4(TDCPTM)(H2O)4, 1 1 1781 857 Cu3(BTC)3(H2O)3 Cu3(TATB)2 19.8 18.4 1925 165 735 1627 10 1 298 146 9.9 1 304 209 9.8 1 298 210 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 4. Continued surface area (m2/g) chemical formulaa Cu2(TCM) Cu(BDC OH)(H2O) common name BET SNU-21H Cu-BDC OH 397 Zn(nbIm)(nIm) ZIF-78 620 Zn(cnIm)(nIm) ZIF-82 1300 Langmuir 695 584 Fe(pz)Ni(CN)4 capacityb (wt %) pressure (bar) temp (K) ref 9.7 9.3 1 1 298 296 206 211 9.3 1 298 212 9.1 1 298 18, 213 9.1 1 298 18, 213 Zn(cyanIm)2 ZIF-96 960 8.8 1.1 298 214 Zn(IDC) IMOF-3 802 8.6 1 298 215 Zn4O(BDC)3 MOF-5, IRMOF-1 2304 2517 2833 1892 3320 Cr(OH)(BDC) Ni3(L-TMTA)2(bpy)4 Zn4O(NO2-BDC)1.19((C3H5O)2-BDC)1.07- MTV-MOF-5-EHI [(C7H7O) 3 BDC]0.74 296 216 298 298 160 217 2784 3.5 1 298 189 3.4 1 298 218 4140 MIL-53(Cr) MIL-47 USO-3-In-A 1 1.2 1.1 1263 Zn2(BDC)2(DABCO) V(IV)O(BDC) In(OH)(BDC) 8.5 4.5 4 3.2 1 298 99, 219 8.5 8.2 1 1 304 298 220 221 8.1 1 296 222 600 930 872 8.1 8 1 1 298 298 189 146 1176 1400 7.7 1.1 298 219 223 Zn4(OH)2(1,2,4-BTC)2 408 7.6 1 295 Pd(μ-H-pymo-N1,N3)2 600 7.3 0.86 293 207 2096 7.3 1.1 298 160 410 7.3 0.9 293 224 1070 7.3 8.6 0.86 1 293 298 225 18, 213, 226 7.2 1 298 205 7.2 1 298 18, 213 6.5 1 298 217 Zn4O(PDC)3 IRMOF-11 Mn(pmdc) 600 950 Pd(2-pymo)2 Zn(cbIm)(nIm), (ZIF-69) Co(tImb) Zn(brbIm)(nIm), (ZIF-81) Zn4O(BTB)2 760 MOF-177 5400 3.6 1 298 48 4508 3.4 1 298 160 3.4 1 298 189 1220 1460 6.8 6.7 1 1 313 313 227 227 6.7 1 298 214 6.7 1 298 18, 213, 226 6.6 1 298 228 6.5 1 298 229 6.4 1 300 230 6.2 1 298 204 6.2 6.2 1 1.1 298 298 231 18, 213 219 Zn8(ade)4(BPDC)6O 3 2TEA Zn8(ade)4(BPDC)6O 3 2TMA TEA@bio-MOF-1 TMA@bio-MOF-1 Zn(almeIm)2 ZIF-93 864 Zn(bIm)(nIm) ZIF-68 1090 Cu(pzdc)2(bpy) CPL-2 633c Zn2(bttb)(dpntcd) YO-MOF 4690 1229 Cu2(pzdc)2(pyz) Al(OH)(bpydc) MOF-253 Zn2(NDC)2(diPyNI) 3 Li Zn(mbIm)(nIm) ZIF-79 Zn4O[(C7H7O)2-BDC]2.49(NO2-BDC)0.51 MTV-MOF-5-EI 2160 2490 840 6.1 1.1 298 Zn2(NDC)2(diPyNI) 5.8 1 298 231 Cu4(O)(OH)2(Me2trzpba)4 5.8 1 298 232 5.7 1 313 227 5.7 1 298 18, 213, 226 5.6 1 297 233 5.5 5.5 1 1 298 298 200 146 Zn8(ade)4(BPDC)6O 3 2TBA Zn(Im)1.13(nIm)0.87 TBA@bio-MOF-1 ZIF-70 en-CuBTTri USO-3-In Zn2(TCPB)(DPG) 3 Li Zn4O(TPDC)3 Cu2(hfbba)2(3-mepy)2 3 2DMF 3 3-mepy 1210 830 1730 1970 231c Cd6(CPOM)3(H2O)6 H3[(Cu4Cl)3(BTTri)8(en)3.75] In(OH)(BDC) 1020 345 930 1912 F-MOF-4 736 376 5.4 1 298 231 5.4 1 298 234 5.3 1 293 235 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 4. Continued surface area (m2/g) chemical formulaa common name BET Zn8(ade)4(BPDC)6O Zn2(bpdc)2(bpe) bio-MOF-1 1680 137.8c Zn4O(BDC-NH2)3 IRMOF-3 2160 Langmuir 5.2 5.2 K2(DABCO-H2)[Zr(ox)4] Zn2(bpy)(TCMO) 1150 Zn2(TCMO) Cu2(TP)3(OH) Zn2(TCPBDA) Cu-TP-1 SNU-30 Cu(dImb) capacityb (wt %) pressure (bar) temp (K) 1 1 313 298 ref 227 236 5.1 1.1 298 160 5 1 313 196 4.7 1 298 189 5.1 1 298 237 5 1 298 238 5 1 298 239 258 707 286 740 5 4.9 1 1 298 298 240 241 435 579 4.7 1 298 242 4.7 1 298 243 Zn2(bpdc)2(bpee) Zn(hymeIm)2 ZIF-25 1110 4.7 1 298 214 Zn4O(BDC-C2H4)3 IRMOF-6 2516 4.6 1.2 298 160 4.6 1 298 231 4.5 1.1 298 214 4.5 4.3 1 1 298 298 244 189 Zn2(TCPB)(DPG) Zn(hymeIm)2 ZIF-97 564 Zn2(bcphfp) Zn(meIm)2 FMOF-2 ZIF-8 378 1135 Cr3O(H2O)2F(BDC)3 MIL-101(Cr) 2674 Cu2(hfbba)2(3-mepy)2 Cu F-MOF-4B 1768 Zn3(OH)(p-CDC)2.5 Zn4O(BDC)(BTB)4/3 UMCM-1 4034 5182 350 Cu(2-pymo)2 Zn20(cbIm)39(OH) ZIF-100 Zn2(hfbba)2(3-mepy)2 3 (3-mepy) Zn(cbIm)2 Zn F-MOF-4B ZIF-95 Cu(etz) MAF-2 ZIF-71 Mg(3,5-PDC) Cu(hfipbb)(H2fipbb)0.5 Mg-MOF-1 Zn2(TCPBDA)(bpta) SNU-31 1 298 245, 246 3.8 1 298 189 3.8 0.86 293 225 298 18, 247 1050 1240 3.7 3.7 1 1 293 298 235 18, 247 3.6 1 298 248 3.4 0.86 293 207 3.4 1 298 249 250 3 1 303 154 3 1 298 251 652 2.8 1 298 214 2.7 2.6 1 1 298 298 252 253 308 1 298 241 1 298 254 2.4 1 298 160 2.4 1 298 254 Pd(μ-I-pymo-N1,N3)2 1.9 0.86 293 207 1.9 1 304 255 1.6 1.4 1 1 298 296 254 222 249 Nd-RPF9 345 2.6 2.6 Co(doborDC)2 Nd6(OH)9(HSCA)(SCA) MOF-2 4.1 1 Co4(OH)2(doborDC)3 Zn3(BDC)3 181 235 3.8 Sc2(BDC)3 Zn(dcIm)2 319 293 780 Cd-ADA-1 [Fe(III)(Tp)(CN)3]2Co(II) 1 1 595 Pd(μ-Br-pymo-N1,N3)2 Cd2(ADA)2(bpy) 4.2 4.1 13.45 Co(doborDC)2(py) Zn2(BDC)2(H2O)2 Mn2(ADA)2(bpy) Mn-ADA-1 1.4 1 298 Ni(DBM)2(bpy) Ni-DBM-BPY 0.9 1 298 256 65 0.8 1 293 225 417.7 0.5 1 297 257 Cu(4-pymo)2 Zn3(Ge(4-carboxyphenyl)4)2 Ge4A-Zn a See List of Abbreviations. b Capacities were estimated from adsorption isotherms in cases where the values were not specifically reported. c Surface area was calculated by CO2 adsorption. enthalpy of CO2 adsorption is frequently expressed as an isosteric heat of adsorption (Q st) as a function of the quantity of CO2 adsorbed. The Q st value is a parameter that describes the average enthalpy of adsorption for an adsorbing gas molecule at a the volume of the adsorbent beds would also be increased due to the lower density of CO2 adsorption. Owing to the presence of a variety of binding environments for CO2 within the pores of metal organic frameworks, the 737 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 5. Lower-Pressure CO2 Adsorption Capacities in Metal Organic Frameworks at 273 K surface area (m2/g) chemical formulaa Cu2(abtc)3 common name BET SNU-5 Langmuir pressure (bar) ref 38.5 1 258 27.2 1 259 25.9 1 260 24.1 1 261 1460 20.6 1 258 19.7 1 262 1260 19.2 1 258 905 892 18.4 17.9 1 1 206 263 Co4(OH)2(p-CDC)3 1080 16.4 1 264 Pd(μ-F-pymo-N1,N3)2 600 14.2 1.2 207 13.8 1 265, 266 Dy(BTC) 2850 capacityb (wt %) 655 Cu2(EBTC)(H2O)2 Cu-EBTC 1852 Al4(OH)2(OCH3)4(BDC-NH2)3 CAU-1 1268 [Zn2(abtc)(DMF)2]3 SNU-4 Zn2(BTetB) 2844 1370 [Cu2(abtc)(DMF)2]3 SNU-50 Cu2(TCM) [In3O(diazDBC)1.5(H2O)3](NO3) SNU-21 Cu(bpy)2(BF4)2 ELM-11 Cu2(bptb) ZIF-20 Zn(Pur)2 Zn2(BTetB)(DMF)2 Ba2(H2O)4[Ln(DSbpyDO)3(H2O)]Cl Cd(DBNBVP)2(ClO4)2 Zn2(BDoborDC)4 1 3 Pd(μ-H-pymo-N ,N )2 Ni(bpy)2(BF4)2 1217 12.6 1 267 800 12.4 1 268 800 12.1 1 262 718c 1096c 12.1 12.1 1 1 269 270 800c 12.1 1 271 600 11.9 1.2 207 11.9 1 265 272 ELM-31 11.6 1 Cu(dhbc)2(bpy) Zn2(TCPB)(DPG) 740 11 1 273 Ni2(bpy)3(NO3)4 10.6 1 274 2910 860 9.9 8.9 1 1 275 101 3010 8.8 1 276 8.7 1 277 8.1 1 278 Cu2(BPnDC)2(bpy), (SNU-6) Cu3(2,7-NDC)3(DEF)(H2O)2 MOP-23 Cd(nim)2 CdIF-9 2590 760 Li(bim)(DABCO)0.5(H2O) 557c Ni(DBNBVP)2Cl2 H[Zn7(μ3 OH)3(DBS)3] UOC-1 Cu2(CNBPDC)2(DMF)2, (MOF-601) In2(OH)2(obb)2 Co4(m OH2)4(MTB)2 Ni2(bpy)3(NO3)4, [Ethanol templated] 354.1 SNU-15 356c BIF-9-Li 1523 649 8.1 1 279 980 7.3 1 101 518.5 7.2 1 280 7 6.9 1 1 281 274 6.6 1 282 6.6 1 283 6.4 1 262 6.3 1 283 6.3 1 267 Zn(pydc)(dma) LiB(4-MIm)4 Zn2(BTetB)(py-CF3)2 1818 390 CuB(4-Mim)4, (BIF-9-Cu) 1287 1524 Cu2[pegBTB)](H2O)2 Mn(NDC)(DEF) 191 Zn4O(BDC)3 Zn6(μ4-O)[bptb]2[H2bptb] (IRMOF-1, MOF-5) Cd(mim)2 CdIF-1 6.2 6.1 2420 MOF-602 Cu2(IBPDC)2(Py)1.67(H2O)0.33 MOF-603 Zn2(BPNDC)2(bpy) SNU-9 Cd(eim)2 [Na6(H2O)6][Eu(DSbpyDO)4] 3 Cl 277 910 5 1 101 460 5 1 101 4.7 1 179 4.6 1 273 4.3 4.3 1 1 286 287 147c 283 2420 426c Er2(PDA)3 738 276 1 1 1030 CdIF-4 285 267 5.6 Cu(dhbpc)(bpy) Ln[(CPI)(H2O)2] Cd(CEbnbpy)(H2O)2 3 (ClO4)2 284 1 1 5.2 Li2(bim)2(DABCO) Cu2(2-MeBPDC)2(DMA)2 6.2 4.0 1 276 3.8 1 288 2.3 1 289 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 5. Continued surface area (m2/g) chemical formulaa common name BET Cu(atz)(benzDHP) Li(bim)(4,40 -bipy) Langmuir capacityb (wt %) 69 2.2 1.8 pressure (bar) ref 1 1 290 277 Co(UO2)6(PO3CH2CO2)3O3(OH)(H2O)2 142c 0.74 1 291 Zn(DHBP)(DMF)2 209 0.56 1 292 a See List of Abbreviations. b Capacities were estimated from adsorption isotherms in cases where the values were not specifically reported. c Surface area was calculated by CO2 adsorption. single-site Langmuir isotherms: specific surface coverage and is usually evaluated using two or more CO2 adsorption isotherms collected at similar temperatures (usually within 10 K of each other). Usually the temperature-dependent isotherms are first fit to a high-order polynomial equation to obtain an expression for the pressure (P) in terms of the quantity of CO2 adsorbed (N), and the Q st values are subsequently computed using the Clausius Clapeyron equation,293 which takes the form ðln PÞN ¼ ðQ st =RÞð1=TÞ þ C N ¼ n 1 m ai N i þ bi N i T i¼0 i¼0 ∑ ∑ ð1Þ ð2Þ where ai and bi are virial coefficients, and m and n are the number of virial coefficients required for adequate fitting of the isotherms. The isosteric heat of adsorption can then be evaluated using the following expression: m Q st ¼ R ∑ ai N i i¼0 ð3Þ and it follows that the zero-coverage isosteric heat of adsorption is given by: Q st ¼ Ra0 ð5Þ where qsat,A and qsat,B are the saturation loadings and bA and bB are the Langmuir parameters for the sites A and B, respectively. An advantage of this type of expression is that it allows a more accurate fitting of the isotherms compared with the virial-type equations, especially when there is a large difference in the adsorption enthalpy of the adsorption sites. Once the isotherms are fit, the isosteric heat of adsorption can then be calculated by reducing eq 5 to a form in which P is expressed as a function of N, followed by the use of eq 1.48 Table 6 lists the zero-coverage isosteric heat of CO2 adsorption for metal organic frameworks reported to date. As expected, at the lowest coverages, where the magnitude of the isosteric heat of adsorption is largely a function of the binding strength of the strongest binding sites within the material, metal organic frameworks bearing amine functionalities or highly polarizing adsorption sites display the highest values. Note that a thorough evaluation of the performance of a given metal organic framework should consider the isosteric heat of adsorption over the entire adsorption range (not just at zerocoverage), which is greatly affected by the pressure regime in which the capture process is to be performed. Furthermore, for some metal organic frameworks where multiple studies have been performed, the values reported for the isosteric heat of adsorption were found to vary significantly. For example, the Q st values at zero-coverage reported for CO2 adsorption within HKUST-1 range from approximately 15 to 35 kJ/mol, which presumably is a result of variations in the synthesis and activation procedures employed for preparation of the samples, and is also likely to be influenced by the method in which the value is calculated. This serves to highlight the importance of ensuring that the experimental conditions are optimized prior to evaluation of the adsorption properties of any metal organic framework. where T is the temperature, R is the universal gas constant, and C is a constant. The Qst at each adsorption level is readily obtained from the slope of the plots of (ln P)N as a function of (1/T). One further parameter of interest for evaluating the strength of CO2 binding, particularly at the lowest CO2 pressures, is the zero-coverage isosteric heat of adsorption. This parameter gives an indication of the strength of the strongest binding sites within the material, which, depending on its magnitude, can subsequently be attributed to certain chemical features of the pore surface, such as exposed metal cation sites or amine functionalities. The zero-coverage isosteric heat of adsorption is evaluated by first fitting the temperature-dependent isotherm data to a virial-type expression,294 which can be written ln P ¼ ln N þ qsat, A bA P qsat, B bB P þ 1 þ bA P 1 þ bB P 2.3. Selectivity for CO2 ð4Þ In CO2 capture applications, a high selectivity for CO2 over the other components of the gas mixture is essential. This selectivity can originate from two main mechanisms. In sizebased selectivity (kinetic separation), a metal organic framework with small pore size may permit molecules only up to a certain kinetic diameter to diffuse into the pores, allowing the molecules to be separated based on size. For CO2/N2 and CO2/ H2 separations, the relatively similar kinetic diameters of the molecules (see Table 2) would require materials operating on a size-selective mechanism to possess very small pores, which may An alternative method for calculating the isosteric heat of adsorption involves the use of a single- or dual-site Langmuirtype expression to first describe the adsorption isotherms at different temperatures. The dual-site model is particularly of benefit when performing the fit for materials featuring a combination of both strong and weak binding sites and assumes that the adsorption occurs with a Langmuir-type behavior at two separate sites.48 Thus, the total quantity adsorbed at a given pressure is given by the sum of two 739 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 6. Zero-Coverage Heat of CO2 Adsorption in Metal Organic Frameworks material chemical formulaa common name functionality type Qst (kJ/mol) ref H3[(Cu4Cl)3(BTTri)8(mmen)12] Cu-BTTri-mmen amines 96 198 H3[(Cu4Cl)3(BTTri)8(en)3.75] Cu-BTTri-en amines 90 200 amines 77 295 Zn(DCTP)(DABCO) Cr3O(H2O)3F(BTC)2 MIL-100(Cr) exposed cations 62b 167 Zn8(ade)4(BPDC)6O 3 TBA TBA@bio-MOF-1 amines 55 227 NH2-MIL-53(Al), USO-1-Al-A amines 50 146 38 147 48 48 197 261 Al(OH)(NH2-BDC) Cu3(TATB)2 Al4(OH)2(OCH3)4(BDC-NH2)3 CuTATB-30 CAU-1 amines Mg2(dobdc) Mg-MOF-74, CPO-27-Mg exposed cations 47 150 41 131 39 190 Co2(ade)2(CO2CH3)2 bio-MOF-11 amines 45 199 Cr3O(H2O)2F(BDC)3 MIL-101(Cr) exposed cations 44b 167 Ni2(dobdc) Ni-MOF-74 CPO-27-Ni exposed cations 42 193 37 131 Zn2(ox)(atz)2 41 201 Pd(μ-F-pymo-N1,N3)2 40 207 Cu(dImb) 39 242 Pd(μ-H-pymo-N1,N3)2 38 207 38b 38 296 236 exposed cations 37 150 Li@CNT@Cu3(BTC)2 exposed cations 37 297 Li@Cu3(BTC)2 exposed cations 36 297 Al4(OH)8(pyromellitate) Zn2(bpdc)2(bpe) MIL-120 Co2(dobdc) Co-MOF-74 CPO-27-Co Zn8(ade)4(BPDC)6O 3 2Me2NH2 bio-MOF-1 35 227 Zn8(ade)4(BPDC)6O 3 2TEA TEA@bio-MOF-1 35 227 Zn8(ade)4(BPDC)6O 3 2TMA Cu3(TATB)2-catenated TMA@bio-MOF-1 CuTATB-60, PCN-6 35 35 227 197 Al(OH)(BDC) MIL-53(Al), 35b 178 30 146 USO-1-Al Cu3(BTC)2 HKUST-1 exposed cations Ni3(pzdc)2(7Hade)2(H2O)4 CNT@Cu3(BTC)2 Zn4O(BDC)3 exposed cations IRMOF-1, MOF-5 35 298 15c 196 35 (at 5.6 wt %) 210 34 34 297 216 17 164 15c 196 Ni2(pbmp) Ni-STA-12 34b 209 Al12O(OH)18(H2O)3(Al2(OH)4)(BTC)6 MIL-96 33b 186 Cr(OH)(BDC) MIL-53(Cr) 32b 178 Er2(PDA)3 30 289 Fe(pz)Ni(CN)4 Al(OH)(bpydc) 3 0.97Cu(BF4)2 30 30 212 204 30 174, 175 Cu3(BTC)2 3 3H2O Cu(hfipbb)(H2fipbb)0.5 HKUST-1 (hydrated) Zn2(bpdc)2(bpee) [In3O(abtc)1.5(H2O)3] Cu(etz) MAF-2 740 30 253 29 243 29 263 27 248 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 6. Continued material chemical formulaa common name functionality type Qst (kJ/mol) ref Fe4O2(btb)8/3 Cu(pzdc)2(bpy) CPL-2 26 26 299 228 Co(dpt24)2 MAF-25 26 300 [Cu(BDC OH)(H2O)] 26 211 [Ni(Bpene)][Ni(CN)4] 25 301 Sc2(BDC)3 25b 250 25 170 25 302 23 23 300 204 Cu4(TDCPTM)(H2O)4 NOTT-140 NH4[Cu3(μ3 OH)(μ3-4-carboxypyrazolato)3] Co(mdpt24)2 Al(OH)(bpydc) MAF-26 MOF-253 Zn2(BDC)2(DABCO) 22 176 [Na6(H2O)6][Eu(DSbpySO)4]Cl 22 288 H3[(Cu4Cl)3(BTTri)8] CuBTTri exposed cations [Ba2(H2O)4][Ln(DSbpyDO)3(H2O)]Cl 21 200 21 269 Ni2(BDC)2(DABCO) 20 176 Zn4(OH)2(1,2,4-BTC)2 20 223 20 19 238 303 Zn2(bpy)(TCM) Zn2(BDC)2(4,40 -bpy) MOF-508b Zn4O(BDC-NH2)3 IRMOF-3 amines [(Tp)Fe(CN)3]2Co Zn4O(BDC)(BTB)4/3 UMCM-1 15 304 19c 196 13 251 12 169 a See List of Abbreviations. Unless otherwise stated, all Q st values are determined by a fitting to the adsorption isotherms, such as the virial or Clausius Clapeyron equation. b Determined experimentally by microcalorimetry. c Determined experimentally by temporal analysis of products (TAP) method. physisorptive mechanisms. For example, in the case of CO2/ N2 separations, the susceptibility of the carbon atom in CO2 to attack by nucleophiles has led to the investigation of materials possessing strong Lewis bases, such as amines. The interaction of CO2 with an amine can result in a C N bond as observed in the aqueous amine solutions, resulting in highly selective adsorption of CO2 over N2. For O2/N2 separations in oxy-fuel combustion, the ability for O2 to participate in electron transfer reactions has led to the investigation of materials constructed from redoxactive metal centers. 2.3.1. Estimation from Single-Component Isotherms. The most basic method for evaluating the adsorptive selectivity for CO2 from a gas mixture is the calculation of a selectivity factor using the experimental single-component gas adsorption isotherms. Here, the selectivity factor is defined as the molar ratio of the adsorption quantities at the relevant partial pressures of the gases. In the case of post-combustion CO2 capture, the partial pressures of CO2 and N2 are 0.15 and 0.75 bar, respectively. Additionally, selectivity factors should be normalized to the composition of the gas mixture as given by the following expression: limit the diffusion of gases throughout the material. While some metal organic frameworks do exhibit pore apertures in this size regime,305 almost all materials that exhibit high surface areas and high adsorption capacities for CO2 possess pore openings that are significantly larger than the sizes of the molecules. Thus, most studies of metal organic frameworks rely on the separation of the molecules based on adsorptive phenomena. The adsorptive selectivity (thermodynamic separation) arises owing to the difference in affinity of the various components of the gas mixture to be adsorbed on the pore surface of the metal organic framework. For selectivity based upon a physisorptive adsorption mechanism, the separation relies on the gas molecules having different physical properties, such as the polarizability or the quadrupole moment, resulting in a higher enthalpy of adsorption of certain molecules over others. For example, for the CO2/N2 separation relevant to post-combustion CO2 capture, the higher polarizability (CO2, 29.1  10 25 cm 3; N2, 17.4  10 25 cm 3) and quadrupole moment (CO2, 13.4  10 40 C 3 m2; N2, 4.7  10 40 C 3 m2) of CO2 compared with N2 results in a higher affinity of the surface of the material for CO2.306 As will be discussed in full detail in section 3, the selectivity can be further enhanced by installing highly charged groups, such as polar organic substituents or exposed metal cation sites, which take greater advantage of this difference in the polarizability of the molecules.10 Alternatively, the adsorptive selectivity can arise due to chemical interactions between certain components of the gas mixture and surface functionalities of the metal organic framework. Functionalities that recognize certain molecules based on their propensity for participating in specific chemistry can result in much higher selectivities than those obtained from purely S¼ q1 =q2 p1 =p2 ð6Þ where S is the selectivity factor, qi represents the quantity adsorbed of component i, and pi represents the partial pressure of component i. Note that the selectivity factor does not take into account the competition of gas molecules for the adsorption sites on the pore surface owing to the fact that it originates from single-component adsorption isotherms and therefore does not 741 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW HKUST-1, which was attributed to the differences in size between CO2 and H2 coupled with the varied pockets and channels present in HKUST-1. Later, it was demonstrated that CO2/H2 separations can be accurately modeled by IAST by substantiating the results with GCMC calculations in MOF177.87 A comparison of the IAST results reported for Mg2(dobdc) to GCMC simulations found that the theory overpredicts CO2/H2 selectivity in metal organic frameworks containing exposed metal cation sites.317 Here, the differences were attributed to the fact that the configurational bias Monte Carlo (CBMC) calculations neglect orbital interactions. The validity of IAST for CH4/H2 separations has been validated in noninterpenetrated318,319 and interpenetrated320 metal organic frameworks. Additionally, it has been demonstrated that, at least at high pressure, IAST and GCMC predict approximately the same CO2/N2 selectivity in HKUST-1.321,322 The quality and fit of the pure component data must be very good for IAST simulations to be accurate. The sensitivity of the results of IAST to the fit of the pure component data has been studied in detail,323 and it has also been highlighted that the issue of obtaining high accuracy fits to the adsorption isotherms can be avoided by using isotherms predicted by GCMC simulations.324 This approach, however, does not take full advantage of the principle benefit of IAST: once a class of metal organic frameworks has been shown to be accurately described by IAST for a given gas mixture, calculus that can be solved analytically can be used to conveniently evaluate mixed-component adsorption from the experimental single-component data. As a result, determining the origin of any deviation between simulations and IAST for specific adsorbate mixtures or adsorbent types is of paramount importance for rendering IAST a routine treatment of pure-component isotherms.325 2.3.3. Gas Mixtures and Breakthrough Experiments. A relatively straightforward way of experimentally evaluating the performance of metal organic frameworks in gas separations is by performing breakthrough experiments. In a typical setup, the gas mixture is flowed through a sample that has been pressed into a pellet or incorporated into a membrane, and the composition of the outgoing gas stream is monitored, usually by gas chromatography or mass spectrometry. As exemplified in Figure 9 for a CO2/N2 gas mixture, before the material bed becomes saturated with adsorbed gas, the downstream composition will consist of virtually pure N2 owing to its much lower affinity for the framework surface. After the bed is saturated with CO2 and breakthrough occurs, the downstream gas composition eventually corresponds to the input mixture. Such experiments have now been performed on metal organic frameworks for a variety of binary gas mixtures, such as CO2/CO,247 CO2/ N2,247,303,304,326 and CO2/CH4,147,190,213,247,268,302 304,327 329 as well as ternary gas mixtures, such as CO2/N2/CH4,303,304 CO2/N2/H2O,326 and CO2/N2/O2.330 The results provide the ultimate laboratory demonstration of the separation performance of a material prior to scale-up. represent the actual selectivity that would result from the dosing of a mixed gas. Nevertheless, it provides a simple point of comparison for evaluating the performance of different metal organic frameworks, and a table of selectivity factors for selected compounds in the context of post-combustion CO2 capture will be presented in section 3. 2.3.2. Ideal Adsorbed Solution Theory (IAST). In practice, it is challenging to measure directly the adsorption selectivity of an adsorbent for gas mixtures, such as those encountered in CO2 capture applications. However, the performance can be conveniently predicted from the single-component adsorption isotherms of the constituents of the mixed gas via modeling techniques, such as ideal adsorbed solution theory (IAST).307 In this method, the isotherms are collected at the same temperatures, and IAST is applied in order to predict the expected selectivity of the material. Here, we briefly describe the most relevant aspects of IAST, and discuss its potential for identifying and evaluating promising CO2 capture adsorbents and the challenges associated with its application. First, the method used to apply IAST for modeling mixed-gas adsorption behavior from single-component isotherms will be outlined. Using a mathematical fitting of the single-component isotherms the mole fraction of each species in the adsorbed phase can be calculated by solving the expression Z P y =x 3 i i 0 ¼ Isotherm fit for component iðPÞ dp P Z P y =x 3 j j 0 Isotherm fit for component jðPÞ dp P ð7Þ where xi and yi are the adsorbed and bulk phase mole fractions of component i, respectively, and P is the total pressure. To determine the amount adsorbed in the mixture, as opposed to the mole fraction, the following equation can be used: 1 ntotal ¼ xj xi 0 þ 0 ni nj ð8Þ where, at a given pressure, ntotal is the total number of moles adsorbed in the mixture and n0i is the amount of pure component i adsorbed per gram of adsorbent. The two main assumptions of IAST are that the components must both mix and behave as ideal gases and that the surface of the sorbent is homogeneous. The accuracy of the theory begins to diminish at very high mixture fractions of the less-adsorbed component, since under those conditions the computation involved in an IAST calculation requires integration of the single-component of the less-adsorbed species up to extremely high pressures.308 The behavior of gases within flexible metal organic frameworks is also not accurately described by IAST in its usual form. It should be noted, however, that a method for applying IAST to flexible metal organic frameworks (or any flexible solid) has recently been developed.309,310 The IAST method has been used to evaluate gas mixtures relevant to post-combustion19,262,311,312 and precombustion313 315 CO2 capture. Additionally, IAST and grand canonical Monte Carlo (GCMC) simulations are often reported together as a method to validate the results of the IAST calculations. For example, it has been shown that IAST compares well with GCMC simulations of CH4/H2 selectivity in MOF-5 and HKUST-1 and CO2/H2 selectivity in MOF-5.316 However, IAST was found to underpredict the CO2/H2 selectivity in 2.4. In Situ Characterization of Adsorbed CO2 In situ methods are an important component of studying the relationship between the structure and chemical features of metal organic frameworks and the observed CO2 adsorption properties. For example, crystallographic characterization of CO2 adsorbed in metal organic frameworks allows direct observation of the location of the CO2 molecules within the structure, facilitating an understanding of the interactions 742 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW within the material giving rise to the adsorptive behavior. Although such experiments usually require elaborate experimental facilities, in situ infrared spectroscopy offers a more convenient probe that can also afford information regarding the CO2 adsorption sites. In the following sections, we describe some recent examples of the use of these techniques in elucidating the adsorption behavior of CO2 within metal organic frameworks. 2.4.1. Structural Observations. The most informative technique for directly obtaining structural information regarding CO2 bound to the surfaces of metal organic frameworks is via X-ray (or neutron) diffraction studies. Here, an activated singlecrystal or powder sample of a metal organic framework is inserted into a cell, such as a glass capillary, which is subsequently evacuated and dosed with a known (usually small) quantity of high-purity CO2 gas.331 Diffraction data are collected, and structural refinements are subsequently performed in order to reveal the location of the dosed CO2 molecules. Note that although such experiments are conceptually quite simple, a number of issues complicate the full elucidation of the CO2 adsorption sites. Indeed, from a practical point of view, absorption of X-rays by the glass capillary, sample integrity upon evacuation, and disorder of the adsorbate molecules within the void spaces of the framework can significantly affect the quality of the refinements. Here, we discuss a number of key studies in which the location of the CO2 adsorption sites within the metal organic framework have been elucidated by diffraction experiments.248,332,333 Although these initial investigations were performed with pure CO2 gas, the future use of a mixture of gases may allow the separation phenomena within metal organic frameworks to be observed directly, enabling a greater understanding of the chemical and structural features that give rise to enhanced CO2 capture performance. One of the earliest examples of an in situ observation of CO2 adsorption was performed on the flexible MIL-53(Cr) system,334 in which chains of octahedral Cr3+ ions are bridged by 1,4benzenedicarboxylate linkers to form diamond-shaped one-dimensional channels. This framework exhibits a breathing behavior upon adsorption of CO2, resulting in a dramatic change in the unit cell parameters. In the evacuated state, and at relatively low pressures of CO2, the pores are narrow (MIL-53LP, see Figure 10). However, above a pressure of 4 bar of CO2, the powder diffraction patterns indicate the evolution of a phase (MIL-53HP) in which the predominant low-angle peaks are shifted to lower angles, indicating an expansion of the pore dimensions. Indeed, refinement of the unit cell parameters of this phase indicate an enlargement of the cell volume by approximately 50%, which is consistent with the increased filling of the pores with guest molecules and a large step observed in the CO2 adsorption isotherm (304 K) at approximately 5 bar. This structural transition is reversible upon lowering of the pressure, and could be cycled without loss of crystallinity. Importantly, these observations established a direct method for the characterization of the framework flexibility of this structure type, which was initially probed via temperature-dependent solid-state NMR spectroscopy for a hydrated sample of MIL-53(Al).335 Figure 9. (top) A schematic of the configuration of an experimental breakthrough setup and (bottom) example of an idealized breakthrough curve for a mixed gas consisting of 20% CO2 in N2. Figure 10. Powder X-ray diffraction patterns as a function of CO2 pressure in MIL-53(Cr) recorded at 293 K. The framework undergoes a reversible structural transition upon increase of the pressure of CO2. Reproduced with permission from ref 334. Copyright 2007 Wiley-VCH. 743 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW More recently, the zinc-based metal organic framework Zn2(atz)2(ox), with a structure comprising two-dimensional layers of Zn2+ ions coordinated by amine-functionalized triazolate ligands pillared by oxalate ligands, was employed in a single-crystal diffraction study in order to observe framework CO2 interactions.201,336 Figure 11 shows the primary binding site for CO2, in which the carbon atom of the CO2 molecule and the nitrogen of the amine are in close contact, suggesting localized adsorption of CO2 at these sites. Here, the C 3 3 3 N distance of 3.152(6) Å and lack of bending in the CO2 molecule indicates that the interaction is physisorptive, with a relatively weak interaction between the amine moiety and the CO2 molecule. It should be noted that such aromatic amines are not strongly basic and will not generally lead to a high CO2 selectivity via chemisorption. High-resolution powder X-ray diffraction data has been used to observe the binding sites and conformation of the CO2 molecules in Ni2(dobdc) (Ni-MOF-74), which possesses exposed Ni2+ metal cation sites on the pore surface (see Figure 12).193 Importantly, the diffraction data following dosing with CO2 reveals that the guest molecules are indeed coordinated to the open metal sites with a bent O C O angle of 162(3)°. The short Ni OCO2 distance of 2.29(2) Å is indicative of a strong interaction owing to the highly polarizing nature of the Ni2+ adsorption sites. The disorder within the structure presents a difficulty in ascertaining the true geometry of the bound CO2, although infrared studies performed on the same compound are indeed consistent with a perturbation of the local geometry from linear to a slightly bent conformation (see section 2.4.2). Elucidation of the binding sites for CO2 within a metal organic framework provides valuable insight into the specific interactions that lead to enhanced adsorption and selectivity properties. The data obtained from structural characterization of metal organic frameworks containing CO2 can be used to study the structural and chemical features of the pore surface that give rise to such properties and can assist with identifying new synthetic targets that may exhibit high performance in model or pilot scale CO2 capture processes. 2.4.2. Infrared Spectroscopy. Infrared spectroscopy has been shown to be a crucial tool for probing the local interactions between CO2 and the surfaces of metal organic frameworks.209,327,334,337 341 The asymmetric stretching (ν3, 2349 cm 1) and bending (ν2, 667 cm 1) modes of CO2 are infrared-active, while the symmetric stretching mode (ν1, 1342 cm 1) is infrared-inactive.342 In cases where strong binding and polarization of the molecule occurs, infrared-active modes can serve as a handle for identifying the nature of the resulting CO2 adduct. Several studies have been performed in which the adsorption of CO2 at exposed metal cation sites has been probed via infrared spectroscopy.167 In the case of Ni2(dobdc) mentioned above, the adsorption of CO2 onto the exposed Ni2+ adsorption sites resulted in an 8 cm 1 red shift of the ν3 band due to electron donation from the oxygen lone pairs on CO2 to the unoccupied Ni2+ orbitals.193 A broad resonance at 2408 cm 1 was assigned to be a combination mode of ν3 + νM O, where νM O is the stretching mode of the Ni2+—OdCdO adduct; from this assignment νM O was determined to be 67 cm 1, in line with the expected value of ∼70 cm 1. The ν2 bending mode was split into a doublet, with an energy separation of 8 cm 1. As previously discussed, refinement against powder X-ray diffraction data of the gas-loaded sample indicated that the adsorbed CO2 adopted a bent configuration, with a large uncertainty in the OdCdO bond angle. The ν2 doublet was presented as evidence for the Figure 11. A portion of the single-crystal structure of Zn2(atz)2(ox) following dosing with CO2 as viewed along the crystallographic a-axis. Yellow, gray, blue, and red spheres represent Zn, C, N, and O atoms, respectively; H atoms are omitted for clarity. Inset shows interaction between the amine group and the CO2 molecule (light gray speres represent H atoms).201,336 Figure 12. A portion of the crystallographic structure of Ni2(dobdc) following dosing with CO2. Green, gray, and red spheres represent Ni, C, and O atoms, respectively; H atoms are omitted for clarity.193 bending of CO2 and reduction of the local symmetry, despite the large geometric uncertainty. In the related compound Mg2(dobdc), the site-specific enthalpy and entropy of adsorption of CO2 at the exposed Mg2+ sites were calculated from a van’t Hoff plot constructed using variable-temperature infrared spectroscopy data.339 The calculated enthalpy of adsorption is 47 kJ/mol, which agrees well with the zero-coverage isosteric heat of adsorption (see section 2.2) obtained from the adsorption isotherms. The negative value for the entropy of adsorption is presumably a result of the efficient packing of the CO2 molecules at the surface of the material, as well as the degrees of freedom being lowered as a result of the strong interaction with the exposed Mg2+ sites. Similar studies performed on the mesoporous MIL-100 and 101 materials have revealed the direct observation of the interaction of CO2 with the exposed Cr3+ cation sites within these materials.128 In MIL-100, a strong ν3 absorption band at 2351 cm 1 was observed upon dosing of CO2 onto the evacuated material. To confirm the assignment of this band as CO2 molecules adsorbed at the Cr3+ sites, the analogous experiment was performed on the solvated material (which would not be expected to possess unsaturated Cr3+ coordination sites), which did not reveal the evolution of such an absorption band. Several 744 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW bands were also observed at 2207, 2200, and 2193 cm 1 and were assigned as weaker Lewis acidic sites within the material. Interestingly, MIL-101 exhibited only a single band at 2196 cm 1, indicating the absence of strong adsorption sites. A more recent study performed on the MIL-100(Fe) compound revealed that a portion of the Fe3+ metal centers in the material could be controllably reduced during the thermal activation of the compound.343 This was attributed to the departure of one of the anionic ligands (F or OH ) from the coordination sphere of the Fe3+ cation (presumably in the form of a radical), resulting in the generation of an exposed Fe2+ adsorption site. The intensity of the vibrational band corresponding to the strong binding sites at approximately 2173 cm 1 grows as a function of the density of Fe2+ ions in the material up to a concentration of approximately 14 mol %, which is approximately 40% of the available binding sites within the material. In addition to probing of the interactions between CO2 and exposed metal cation sites, the resolution of interactions between CO2 and functional groups on the organic ligands has also been studied via vibrational spectra. Diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) is an effective tool for analyzing the surface chemistry of porous solids, especially when the in situ cell environment can be controlled. As discussed below, this technique has been applied to a handful of functionalized metal organic frameworks to date,344 346 revealing valuable information regarding the nature of CO2 adsorption within the material. The chemical adsorption of CO2 onto basic alkylamines was recently reported in the N,N0 -dimethylethylenediamine (mmen) functionalized framework H3[(Cu4Cl)3(BTTri)8(mmen)12] (mmen-Cu-BTTri), which will be discussed in detail in the context of post-combustion CO2 capture in section 3.198 The N H stretch of mmen was followed as a 5% CO2 in helium mixture was introduced into a high-pressure DRIFTS cell. As shown in Figure 13, the N H stretching band of mmen at 3283 cm 1 disappears upon introduction of CO2 into the cell. Regeneration by heating and vacuum restores the band to the spectrum. Chemisorption, suggested by the large isosteric heat of adsorption measured at zero coverage, was thus confirmed by infrared characterization. The presence of broad, lower energy bands, including the emergence of a band at 1669 cm 1, was further suggestive of the formation of zwitterionic carbamates or carbamic acid in the absence of water. The adsorption of CO2 onto aromatic amines (e.g., aniline derivatives) has also been a subject of in situ studies using DRIFTS.346 Although the nitrogen atom within 2-aminoterepthalic acid (NH2-BDC) is significantly less basic than that of mmen, the formation of Lewis acid base pairs is possible. The adsorption of CO2 onto Zn4O(NH2-BDC)3 (IRMOF-3) and Al(OH)(NH2-BDC) (NH2-MIL-53(Al)) was first reported to result in splitting of the ν2 band through a loss of degeneracy of the in-plane and out-of plane bends.345 This was found for both compounds to result in the formation of electron donor acceptor complexes, as further substantiated by the appearance of additional bands between 1000 and 1500 cm 1 in the case of NH2-MIL-53(Al). Additional work, however, has shown that the mechanism of CO2 adsorption in NH2-MIL-53(Al) is significantly more complicated than originally thought. Density functional theory (DFT) calculations augmented additional DRIFTS measurements to elucidate a surprising flexion-based adsorption mechanism that does not involve direct interaction of CO2 and amines. Strong Lewis acid base interactions would be Figure 13. Infrared spectra collected for a sample of H3[(Cu4Cl)3(BTTri)8(mmen)12] (purple). Upon dosing of the sample with CO2 (red, then green), the N H stretching band of mmen at 3283 cm 1 disappears but is restored following regeneration of the material (blue). Reproduced with permission from ref 198. Copyright 2011 The Royal Society of Chemistry. expected to significantly alter the N H stretching frequencies and intensities of amines, as observed within mmen-CuBTTri. Thus, the negligible distortion of the N H symmetric and asymmetric modes of IRMOF-3 and NH2-MIL-53(Al) is a good indicator that electron transfer between the amine and CO2 is only a minor component of the adsorption energy in those compounds. 2.5. Computational Modeling of CO2 Capture Computational studies relating to the adsorption of CO2 within metal organic frameworks189,234,347 351 are an important endeavor in supporting the experimental work performed in the area. In this regard, grand canonical Monte Carlo (GCMC) simulations have been widely employed in predicting the singleand mixed-component adsorption isotherms in metal organic frameworks.352 354 Furthermore, a number of DFT studies have emerged recently,355,356 with many focused on determining the packing arrangement of CO2 molecules within the pores and the associated enthalpy of adsorption,357 and the perturbation of these as a result of the introduction of certain functionalities to the periphery of the organic ligands.358,359 The latter is aimed at facilitating direct correlations between the electronic structure of the metal organic framework and the observed adsorptive behavior.360,361 Although there has been a significant amount of recent work performed in terms of developing theoretical methods for describing adsorption within metal organic frameworks, significant challenges still exist in the area. These include the treatment of long-range interactions and adsorption of molecules at exposed metal cation sites,362 the applicability of these methods to large-pore systems234,363 (such as MIL-100 and MIL-101), and the development of more sophisticated algorithms that can account for flexibility within materials.364,365 Nevertheless, computational modeling has been recently utilized in the virtual screening for metal organic frameworks with high adsorption capacities,189 new hypothetical structures with high performance parameters,358 and predicting the stability of existing materials toward contaminants, such as H2O.155 While we do not provide a thorough coverage of the theoretical studies pertaining to CO2 adsorption in metal organic frameworks here, we introduce certain computational results as appropriate in the forthcoming sections. 745 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW In Table 7, the materials displaying the highest selectivities are generally those bearing functionalized pore surfaces. Surface functionalities that interact strongly with CO2 (and to a lesser extent N2) frequently increase adsorbent capacity at low pressures. Ideally, for high adsorption selectivities, the CO2 adsorption should be maximized at pressures near 0.15 bar. Note that for metal organic frameworks with strongly polarizing sites, the selectivity values listed in Table 7 likely underestimate the true adsorptive selectivity of the adsorbent. This is because in a realistic flue gas mixture, the strongest binding sites would be predominantly occupied by CO2 owing to its greater polarizability and quadrupole moment. Thus, single-component isotherms overestimate the adsorption of N2 and consequently reduce the calculated selectivity. Utilization of the IAST method (see section 2.3.2) for selectivity calculations minimizes this effect and enables more accurate selectivity values to be calculated.317 Since using IAST to predict selectivity for a gas mixture requires accurate fits of the pure-component isotherms, selectivities are instead reported in Table 7 using the adsorption values obtained directly from the pure-component isotherms. Indeed, IAST values at the appropriate pressures have been reported for only a handful of frameworks, but values for certain adsorbents are mentioned in the discussion that follows. It should be noted that the relevance of selectivity values calculated for flexible frameworks via the aforementioned method has not been fully established, although it is likely that the quantities calculated by this methodology would overestimate the true adsorptive selectivity. This is because flexible frameworks generally adsorb little N2 at 0.75 bar because the apertures of the framework are not opened by N2 in singlecomponent gas adsorption experiments. However, if CO2 affords the necessary gate opening pressure, the quantity of N2 adsorbed is likely to be significantly higher than that observed during the single-component experiment. Since IAST cannot account for such effects, direct selectivity measurements of binary mixtures are likely the only accurate gauge of selectivity in flexible metal organic frameworks. Thus, Table 7 omits selectivity values for flexible frameworks, although singlecomponent CO2 and N2 capacities at the relevant pressures are tabulated. 3. POST-COMBUSTION CAPTURE The combustion of coal in air generates flue gas with a relatively low CO2 concentration (15 16%), while the bulk of the effluent is composed of N2 and other minor components, such as H2O, O2, CO, NOx, and SOx (see Table 1).200 The gas stream is released at a total pressure of approximately 1 bar. Since SOx removal would precede CO2 capture, the flue gas would be expected to enter the CO2 scrubber at temperatures between 40 and 60 °C.366,367 As outlined in section 1.4, an ideal adsorbent for capturing CO2 from post-combustion flue gas would exhibit a high selectivity for CO2 over the other flue gas components, high gravimetric and volumetric CO2 adsorption capacities, minimal energy penalty for regeneration, long-term stability under the operating conditions, and rapid diffusion of the gas through the adsorbent material. The preparation of next-generation metal organic frameworks that satisfy all of these requirements is currently a difficult synthetic challenge, although as discussed below, significant progress has been made in recent years. 3.1. Metal Organic Frameworks for CO2/N2 Separation The lower-pressure (<1.2 bar) CO2 uptake capacities for all metal organic frameworks reported to date are listed in Tables 4 (298 319 K) and 5 (273 K). Many of the frameworks exhibit large CO2 adsorption capacities at pressures at and above 1 bar, owing to their high surface areas. However, these compounds are generally not well suited for post-combustion capture, since the adsorption capacity at lower pressures is a more relevant consideration due to the low partial pressure of CO2.48 Note that in these tables, the adsorption capacities at 1 bar (the most commonly reported value) are tabulated to facilitate a general comparison among different metal organic frameworks. However, the adsorption capacities for CO2 (0.15 bar) and N2 (0.75 bar) at pressures relevant to post-combustion CO2 capture are often not directly reported, and these have been carefully estimated from the literature gas adsorption isotherms and are listed in Table 7. The selectivity calculation for CO2 over N2 is best performed using the adsorption capacities at pressures of approximately 0.15 bar for CO2 and 0.75 bar for N2. Since the total pressure of a flue gas is approximately 1 bar, selectivity calculations based upon the quantity of both CO2 and N2 adsorbed at 1 bar drastically overestimate the fraction of CO2 in post-combustion flue gas and the total pressure of the gas. Table 7 lists also the calculated selectivity values for CO2 over N2 at 298 K for selected metal organic frameworks, as calculated using the molar ratio of the CO2 uptake at 0.15 bar and the N2 uptake at 0.75 bar (eq 6). It is important to emphasize that these selectivity factors are based on pure-component adsorption isotherms and do not necessarily represent the true selectivity of the material in a CO2/ N2 mixture. As such, the direct measurement of multicomponent isotherms, which has been recently performed for CO2/CH4 mixtures,327 is necessary in order to evaluate the accuracy of selectivity factors predicted from single-component isotherms and IAST. To the best of our knowledge, there are no experimentally reported CO2/N2 binary isotherms yet reported. As a result, calculated selectivity factors are useful for preliminary evaluations of different materials. Note that, although gas adsorption measurements have most commonly been made at or below 298 K, it would be of benefit for a more realistic evaluation of post-combustion CO2 capture performance if researchers were to begin reporting adsorption data in the range of 313 333 K. 3.2. Enhancing CO2/N2 Selectivity via Surface Functionalization Tuning the affinity of the framework functionalities toward CO2 is crucial for optimization of the adsorptive properties. In the ideal case, the appropriate pore surface properties would give rise to an adsorbent with high adsorption selectivity and capacity yet minimize the regeneration energy. In the following sections, we outline the various types of functionalities that have been explored to date for preparing materials with enhanced CO2 capture performance, including amines, strongly polarizing organic functionalities, and exposed metal cation sites. 3.2.1. Pores Functionalized by Nitrogen Bases. Metal organic frameworks functionalized with basic nitrogen-containing organic groups have been intensively studied for their CO2 adsorption properties. The dispersion and electrostatic forces resulting from the interaction of the quadrupole moment of CO2 with localized dipoles generated by heteroatom incorporation are typically responsible for the enhanced CO2 adsorption. In some cases, acid base type interactions between the lone-pair of nitrogen and CO2 have been observed. The degree to which 746 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 7. CO2 and N2 Uptake in Selected Metal Organic Frameworks at Pressures Relevant to Post-combustion CO2 Capture material chemical formulaa Mg2(dobdc) Ni2(dobdc) common names CO2 uptake at N2 uptake at 0.15 bar (wt %)b 0.75 bar (wt %)b Mg-MOF-74, Mg-CPO-27 Ni-MOF-74 selectivityc temp (K) ref 20.6 1.83 44 303 48 18.9 1.40 52.3 313 48 16.7 1.08 58.8 323 48 14.5 0.87 61.1 333 48 16.9 2.14 30 298 131, 189 298 189 CPO-27-Ni Co2(dobdc) Co-MOF-74 CPO-27-Co 14.2 Cu3(BTC)2 HKUST-1 11.6 0.41 101 293 194 H3[(Cu4Cl)3(BTTri)8(mmen)12] mmen-Cu-BTTri 9.5 0.20 165 298 198 8.3 293 201 Zn-MOF-74 7.6 296 150 Zn2(ox)(atz)2 Zn2(dobdc) CPO-27-Zn Pd(μ-F-pymo-N1,N3)2 293 207 Cu3(TATB)2 Co2(adenine)2(CO2CH3)2 CuTATB-60 bio-MOF-11 5.8 5.4 0.82 0.28 24 65 298 298 197 368 Fe3[(Fe4Cl)3(BTT)8(MeOH)4]2 Fe-BTT 5.3 0.95 18 298 140 4.0 0.34 39 298 204 0.36 30 298 18, 213 298 146 Al(OH)(bpydc) 3 0.97Cu(BF4)2 6.5 Zn(nbIm)(nIm) ZIF-78 3.3 Al(OH)(2-amino-BDC) NH2-MIL-53(Al), USO-1-Al-A 3.1 H3[(Cu4Cl)3(BTTri)8] Cu-BTTri 2.9 Cu2(bdcppi)(DMF)2 SNU-50 2.9 H3[(Cu4Cl)3(BTTri)8(en)3.75] Zn2(bpdc)2(bpee) en-Cu-BTTri 2.3 2.1 0.49 0.17 0.02 19 44 d 298 200 298 202 298 298 200 243 Ni2(2-amino-BDC)2(DABCO) USO-2-Ni-A 2.1 298 146 Cu3(BPT(N2))2 UMC-150(N)2 1.9 298 189, 369 Cu3(BPT)2 UMCM-150 1.8 298 189, 369 298 262 298 146 Zn2(BTetB) Al(OH)(BDC) 1.8 MIL-53(Al), USO-1-A 19 1.7 Zn2(bmbdc)2(4,40 -bpy) Ni2(BDC)2(DABCO) V(IV)O(BDC) 0.31 1.4 USO-2-Ni MIL-47 0.01 d 1.2 1.1 298 370 298 298 146 189 Al(OH)(bpydc) MOF-253 1.0 0.37 9 298 204 Zn20(cbIm)39(OH) ZIF-100 1.0 0.15 22 298 247 Zn4O(NO2-BDC)1.19((C3H5O)2-BDC)1.07- MTV-MOF-5-EHI 1.0 298 219 Zn(MeIM)2 ZIF-8 0.6 Zn4O(BDC-NH2)3 Zn4O(BTB)2 IRMOF-3 MOF-177 0.6 0.6 Zn4O(BDC)(BTB)4/3 UMCM-1 Zn4O(BDC)3 MOF-5, IRMOF-1 ((C7H7O)2-BDC)0.74 Zn2(BTetB)(py-CF3)2 0.9 Co4(OH)2(doborDC)3 298 262 298 189 298 298 160, 189 48 0.5 298 189 0.5 298 189 298 254 0.5 0.06 0.39 0.08 50 4 18 a See List of Abbreviatons. b When not directly reported, values were estimated from adsorption isotherms in the corresponding reference. c Selectivities were calculated from the pure-component isotherms by dividing the mass of CO2 adsorbed at 0.15 bar by the mass of N2 adsorbed at 0.75 bar according to eq 6 in section 2.3.1. d Single-component isotherms are not expected to be meaningful predictors of mixed-component selectivity for these flexible materials. Perhaps owing to the ease of synthesis, a particularly large number of metal organic frameworks with organic bridging units derived from nitrogen-containing heterocycles have been studied in the context of CO2 capture.140,200,202,205,215,224,228,230,232,235,241,248,251,268,269,277,288,329 However, the incorporation of these heterocycles generally improves capacity only modestly at low pressures. Among the nitrogen incorporation enhances CO2 adsorption depends significantly on the nature of the functional group. To date, three major classes of nitrogen-functionalized metal organic frameworks have been synthesized: heterocycle (i.e., pyridine) derivatives, aromatic amine (i.e., aniline) derivatives, and alkylamine (i.e., ethylenediamine) bearing frameworks. 747 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 8. Select pKa Values for the Dissociation of Protonated Amine Complexes in Water373 parent amine pKa aniline 4.60 pyridine benzylamine 5.23 9.38 ethylamine 10.79 the enhanced adsorption. It was recently shown for the wellstudied compound Al(OH)(NH 2 -BDC) (NH2 -MIL-53(Al)) that the aromatic amine has little direct interaction with adsorbed CO2 at low pressures.346 However, the presence of the amine favors the formation of a flexed, narrow-pore structure upon evacuation. This same narrow-pore structure, in which hydrogen bonding is observed between the aromatic substituents and the Al-based clusters, is not observed for the parent MIL53(Al) framework until the material is cooled to temperatures below 150 K. In the narrow-pore structure, CO2 molecules can interact with multiple pore surfaces simultaneously. Furthermore, DFT calculations showed that interactions between CO2 and hydroxyl groups that line the pore surfaces were stronger in the amino functionalized material because of the increased acidity of the hydroxyl moieties. While enhanced adsorption of CO2 in NH2-MIL-53(Al) was originally attributed to direct interactions between the amino functionality and CO2,147 in situ infrared spectroscopy experiments showed that the N H resonances of NH2-MIL-53(Al) were only slightly perturbed by CO2 adsorption. Amines are generally expected to enhance CO2 adsorption in porous materials by acid base chemistry, electrostatic forces, or enhanced dispersion forces. In this case, however, the actual mechanism of binding, when fully elucidated, was shown to be quite different from those observed in other frameworks. For heterocycle- and aromatic amine-based frameworks, enhanced adsorption at low pressures is likely only partially attributable to the basicity of the nitrogen donor atom. The pKa values of ammonium complexes can help explain the basicity differences between different classes of amines. Table 8 lists the pKa values for selected ammonium complexes. Note that the conjugate acids of pyridine and aniline are significantly more acidic than the conjugate acids of benzylamine or ethylamine. The presence of electron-withdrawing carboxylate functionalities ortho or para to the amine would be expected to further reduce the basicity of aniline derivatives. Thus, the enhanced adsorption of CO2in metal organic frameworks containing pyridine and aniline derivatives is likely primarily attributable to the increased number of polarizing sites that decorate the pore walls. Without the formation of Lewis acid base pairs, the binding of CO2 is still physisorptive and regeneration conditions are generally very mild. In order to mimic the chemisorptive interactions that are observed in aqueous amine scrubbers, more basic amine species need to be incorporated onto the pore surfaces of metal organic frameworks. Very strong adsorption of CO2 has indeed been shown to be possible via incorporation of alkylamines into metal organic framework pores. As illustrated in Figure 15, alkylamine incorporation onto the open metal sites of CuBTTri was found to be an effective method for postsynthetically modifying this metal organic framework to enhance the CO2 binding.198,200 Note that although not explored initially for CO2 adsorption, the grafting of amine moieties onto exposed Figure 14. A portion of the crystal structure of Co2(adenine)2(CH3CO2)2 (bio-MOF-11).368 Turquoise, gray, blue, and red spheres represent Co, C, N, and O atoms, respectively; H atoms are omitted for clarity. The turquoise tetrahedra represent the coordination sphere of the Co2+ ions. best-performing metal organic frameworks featuring nitrogencontaining heterocycles are those incorporating biologically relevant moieties.210,227 The ratio of heteroatoms to carbon atoms in these frameworks can be quite large, resulting in pore surfaces with significant surface polarization. For example, bioMOF-11 was synthesized from adeninate and acetate linkers (see Figure 14).368 Its CO2 adsorption capacity is ca. 5.8 wt % at 298 K and 0.15 bar, with a corresponding isosteric heat of adsorption at zero-coverage of 45 kJ/mol. The large initial isosteric heat is likely partially attributable to the presence of an aromatic amine that also decorates the pore surface. However, the effects of the nitrogen heterocycle are considerable because the CO2 capacity of bio-MOF-11 exceeds that of any other aromatic amine functionalized framework reported to date at 1 bar, despite a surface area of only 1040 m2/g. The commercial availability of aromatic amine containing linkers, especially 2-aminoterephthalic acid (NH2-BDC), and the expected affinity of amino groups toward CO2 has generated significant interest in aromatic amine functionalized frameworks.146,201,258,263,290,371,372 In particular, the IRMOF series of frameworks provides a basis for elucidating the effects of aromatic amines within metal organic frameworks. At 298 K and 1.1 bar, Zn4O(BDC)3 (IRMOF-1 or MOF-5) adsorbs approximately 4.6 wt % CO2, while the amine-functionalized variant Zn4O(NH2-BDC)3 (IRMOF-3) adsorbs 5.0 wt % CO2, despite a decrease in the BET surface area from 2833 to 2160 m2/g.160 The reduced surface area of IRMOF-3 decreases the high-pressure CO2 adsorption capacity, but the higher uptake observed within the material at lower pressures leads to it being a higherperformance material compared with IRMOF-1 in the context of post-combustion CO2 capture. We note that there has been a report of significantly greater adsorption of CO2 in IRMOF-1 (8.5 wt % at 1 bar).216 However, this anomalous result stands in contrast to the lower value more frequently reported and has yet to be reproduced to the best of our knowledge. Amine functionalization has been shown to enhance CO2 capacity in a number of other metal organic frameworks, including Ni2(NH2-BDC)2(DABCO), Al(OH)(NH2-BDC) (NH2MIL-53(Al)) and In(OH)(NH2-BDC), when their low-pressure capacities are compared with that of the parent material.146 However, the amine may not always be directly responsible for 748 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Figure 15. Functionalization of the metal organic framework CuBTTri through binding N,N0 -dimethylethylenediamine (mmen) to the open metal coordination sites; dark red, green, gray, and blue spheres represent Cu, Cl, C and N atoms, respectively. The basic amine groups dangling within the framework pores lead to strong CO2 adsorption, with a zero-coverage isosteric heat of adsorption of 96 kJ/mol. Reproduced with permission from ref 198. Copyright 2011 The Royal Society of Chemistry. Figure 16. (top) CO2 (9) and N2 (b) adsorption isotherms collected at 298 K for mmen-Cu-BTTri (green symbols) and Cu-BTTri (blue symbols) and (bottom) a plot of the isosteric heat of adsorption of CO2 (9) and N2 (b) in mmen-Cu-BTTri (green symbols) and Cu-BTTri (blue symbols). Reproduced with permission from ref 198. Copyright 2011 The Royal Society of Chemistry. metal cations was also previously demonstrated in the MIL-101 system,374 indicating that this approach may indeed be applicable to a wide range of materials with pores that are furnished with such sites. The compound H3[(Cu4Cl)3(BTTri)8(mmen)12] (mmenCu-BTTri) was synthesized by stoichiometric functionalization of Cu-BTTri with the sterically encumbered secondary amine N,N0 -dimethylethylenediamine (mmen).198 A 3.5-fold increase in gravimetric capacity at 0.15 bar (to ∼9.5 wt %) compared with the nonfunctionalized Cu-BTTri framework replete with open metal sites (see Figure 16, upper) was achieved. The highly selective chemisorptive process in this framework strongly affects only CO2 adsorption. Indeed, N2 adsorption decreased at all pressures compared with the parent framework, because of the significantly reduced surface area resulting from the pore modification. Thus, the IAST selectivity of mmen-Cu-BTTri for CO2 at conditions relevant to post-combustion carbon capture is 327 at 298 K. This high selectivity is associated with a large isosteric heat of CO2 adsorption at low pressures, approaching 96 kJ/ mol at zero-coverage, which is substantially higher than the ca. 15 kJ/mol observed over the entire sorption range for N2 (see Figure 16, lower). In the case of CO2, the isosteric heat initially sharply decreases with CO2 adsorbed, which is consistent with the highest-affinity amine sites becoming occupied with CO2 molecules. However, despite this decrease, the magnitude is significantly higher than that observed for the parent framework, which exhibits an almost constant isosteric heat of approximately 23 kJ/mol. Although the large initial adsorption enthalpy in mmen-Cu-BTTri suggests that high temperatures would be required for desorption of the CO2 from the amines, the regeneration conditions were shown to be mild. Following activation, the material adsorbed nearly 7 wt % CO2 from a 15% CO2 in N2 mixture. Complete regeneration was accomplished with a N2 purge and fast temperature ramp to 60 °C, with no capacity loss observed over the course of 72 cycles (27 min per cycle). More detailed work, such as optimization of the amine, and the consideration of the effect of the temperature increase expected from the large quantity of heat originating from the high isosteric heat of adsorption, and the influence of the other minor impurities of the post-combustion flue gas on the performance of the material, would allow the promise of this approach for industrial applications to be further evaluated. 3.2.2. Other Strongly Polarizing Organic Functional Groups. In addition to the amine and other nitrogen-based functionalities discussed in the previous sections, organic linkers with heteroatom functional groups (other than amines) have also been investigated in detail for their effects on the CO2 adsorption behavior.18,101,179,206,211,213,214,225,226,231,233,235,238,239,244,253,255, 262,267,270,272,273,275,276,278 280,286,287,292,347 These functional groups include hydroxy, nitro, cyano, thio, and halide groups, and the degree to which CO2 adsorption is enhanced in these cases depends largely upon the extent of ligand functionalization and the polarizing strength of the functional group. In general, more strongly polarizing groups will influence CO2 adsorption more favorably. A systematic understanding of polarization strength is perhaps best observed via the large number of functionalized zeolitic imidazolate frameworks that have been reported to date.18,213,226 749 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW This subclass of metal organic frameworks is convenient for understanding the effects of different functional groups, because all frameworks incorporate imidazolate-based organic linkers and most have tetrahedral Zn2+ cation nodes. While there are important topological differences between the many frameworks, a wide number of isostructural compounds with different functional groups have also been reported, allowing comparisons to be made between the compounds. For example, a number of mixed-linker frameworks of the GME topology were studied for functional group and pore size effects on adsorption.213 Linkers containing various functional groups afforded a series of frameworks with drastically different pore sizes. Adsorption was found not to correlate with the pore size, which varied between 7.1 and 15.9 Å. Different functional groups, however, appear to impact volumetric adsorption capacity. The compound Zn(nbIm)(nIm) (ZIF-78) exhibits the largest volumetric capacity because every imidazolate and benzimidazolate linker was modified with a polarizing nitro group. Frameworks synthesized from cyanoand nitro-functionalized (ZIF-82), chloro- and nitro-functionalized (ZIF-69), and bromo- and nitro-functionalized (ZIF-81) imidazolate linkers performed similarly to each other but slightly worse than ZIF-78. Frameworks partially functionalized with alkyl groups adsorbed the least quantity of CO2 on a volumetric basis. As expected, the greater the surface polarization engendered by the functional group in the framework, the higher the CO uptake capacity is at low pressures. The foregoing conclusion is not immediately apparent if framework performance is only benchmarked by gravimetric capacity. On a gravimetric basis, less-functionalized, lighter frameworks outperformed the more functionalized, heavier materials. The compound ZIF-70, composed of nonfunctionalized and nitro-functionalized imidazolate linkers (see Figure 17), exhibits a higher gravimetric capacity than ZIF-78. For the latter materials, the presence of benzimidazolate linkers increases the molecular weight significantly, but with the same connectivity, the unit cell volumes of ZIF-78 (11 514 Å3)213 and ZIF-70 (11 386 Å3)226 are very similar. Thus, the best comparison between ZIF-70 and ZIF-78 is the quantity of CO2 adsorbed within a fixed volume. Assuming void spaces between crystallites are similar, a particular volume of ZIF-78 would adsorb 33% more CO2 than the same volume of ZIF-70, which for a stationary application like carbon capture is relevant to infrastructure costs. For isostructural materials, volumetric capacity calculations can elucidate the effects of functional groups without discriminating against the additional mass. The IRMOF series of metal organic frameworks are similarly suitable for investigating functional group effects on CO2 adsorption in a controlled manner. In one study, multiple functional groups from different types of organic linkers were incorporated into individual IRMOF crystallites.219 Volumetric CO2 adsorption capacities were again utilized to account for mass differences between frameworks. Two potential benefits of this mixed linker technique are apparent. First, mixed ligand frameworks can incorporate functional groups (like 2-nitroterephthalic acid) into the IRMOF structure that could not be incorporated when used alone. In particular, the nitro functionality was a common feature among the best performing frameworks synthesized. Second, the presence of multiple functional groups along a single pore surface can afford properties that exceed those expected from a simple linear combination of the individual components. This was demonstrated by the synthesis of Zn4O(NO2-BDC)1.19((C3H5O)2-BDC)1.07((C7H7O)2-BDC)0.74 (MTV-MOF-5-EHI) (see Figure 18), Figure 17. A portion of the crystal structure of Zn(Im)(nIm) (ZIF-70) as viewed along the [001] direction. Yellow, gray, blue, and red spheres represent Zn, C, N, and O, respectively; H atoms are omitted for clarity.226 Figure 18. MTV-MOF-5-EHI is composed of the three ligands shown.219 The incorporation of multiple ligands within a single crystallite was reported to improve CO2 adsorption capacity at low pressures. which adsorbs 7.7 wt % CO2 at 1 bar and has a CO2/CO selectivity at 298 K that is over 4 times greater than that observed within MOF-5, which does not possess any substitutions on the organic backbone. While the performance and stability of the IRMOF series of metal organic frameworks may ultimately make them unsuitable for CO2 capture applications, mixed ligand and mixed functionality frameworks could potentially lead to materials with finely tunable properties. Theoretical work has also modeled the effects of certain functional groups on CO2 adsorption.360 For example, a series of functionalized zirconium metal organic frameworks based upon the UiO-66 structure-type with substituted BDC2 ligands were constructed via DFT calculations.375 GCMC simulations were used to predict isotherms for CO2 and CH4. The authors concluded that strongly polarizing groups enhanced CO2 adsorption through increased isosteric heats of adsorption. Futhermore, it was observed that confinement effects attributable to reduced pore volumes could further enhance CO2 adsorption. Among the most promising materials of this class, the sulfonic acid and carboxylic acid functionalized materials displayed the largest selectivity enhancement in the separation of CO2 from CH4.375 3.2.3. Exposed Metal Cation Sites. Another strategy that has been explored as a means for improving the affinity and selectivity of metal organic frameworks toward CO2 over N2 is the generation of structure types bearing exposed metal cation sites on the pore surface.128 132 These sites are usually obtained following desolvation of the material, where one of the solvent molecules in the coordination sphere of the metal center is removed in vacuo at elevated temperatures. Such metal binding 750 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW sites have been previously shown to facilitate close approach of guest molecules to the pore surface, which is of benefit for raising the adsorption enthalpy and storage density of gases, such as H2 and CH4,13,376,377 and in Lewis acid catalysis.378 In the context of post-combustion CO2 capture, the open metal cation sites serve as charge-dense binding sites for CO2, which is adsorbed more strongly at these sites owing to its greater quadrupole moment and polarizability compared with N2. The earliest studies of metal organic frameworks possessing exposed metal cation sites were performed predominantly on Cu3(BTC)2 (HKUST-1, Figure 7).133,337,379 Here, the solvent molecules present on the axial sites of the paddlewheel units can be removed to afford open Cu2+ adsorption sites. These sites interact more strongly with CO2 due to the high charge density of the Cu2+ cation, resulting in a zero-coverage isosteric heat of adsorption of 29.2 kJ/mol. A number of independent studies have reported the adsorption isotherms for this compound, wherein adsorption capacities range from 15.0 to 18.4 wt % at 1 bar and 298 K.160,189,191,192,194 The difference in these values likely stems from the degree of purity of the material or the degree of activation (desolvation) of the compound. Note that the adsorption properties of metal organic frameworks containing bound solvent molecules are highly dependent upon the desolvation conditions. Furthermore, careful handling of the materials following activation is essential, as a result of the propensity of the exposed metal cation sites to become hydrated as a result of even brief exposure to atmospheric moisture.115 The adsorption of CO2 within the chromium-based metal organic frameworks Cr3O(H2O)3F(BTC)2 (MIL-100) and Cr3O(H2O)2F(BDC)3 (MIL-101) has been studied.167 These materials exhibit BET surface areas of 1900 and 4230 m2/g, respectively, and feature exposed Cr3+ sites following removal of the H2O molecules originally bound to the metal centers.128 Accordingly, the high charge density of the metal ion affords a zero-coverage isosteric heat of adsorption of 62 kJ/mol in the case of MIL-100 and 44 kJ/mol for MIL-101. The storage capacities were probed at high pressures, and CO2 uptakes of over 40 wt % were observed for both compounds at 50 bar and 304 K. Although no low-pressure (<1.0 bar) data were reported, one advantage of these materials is their high chemical and thermal stability. Indeed, both compounds are water-stable and can be heated to temperatures approaching 300 °C without degradation of the framework structure, a crucial characteristic for post-combustion CO2 capture where significant quantities of H2O are present in the flue gas. The M2(dobdc) structure type represents one of the most wellstudied families of materials with regard to metal organic frameworks bearing exposed metal cation sites (see Figure 8).150,380 384 The high density of binding sites decorating the hexagonal onedimensional pores endow the materials with high adsorption capacities for CO2 at 1 bar and 296 K, ranging from 19.8 to 26.0 wt % within Zn2(dobdc) and Mg2(dobdc), respectively.150 In fact, the value for Mg2(dobdc) represents the highest low-pressure gravimetric and volumetric adsorption capacity for CO2 in a metal organic framework, despite its relatively low surface area (SABET = 1495 m2/g), demonstrating the importance of furnishing the pores with a large number of high-affinity binding sites. The zero-coverage isosteric heat of CO2 adsorption across this series is significantly affected by the metal cation, wherein Mg2(dobdc) was observed to display the highest affinity (Q st = 42 kJ/mol), while Zn2(dobdc) exhibited the weakest interactions (Q st = 26 kJ/mol) among the compounds studied. The isosteric heat of adsorption plot Figure 19. Isosteric heat of adsorption, Qst, as a function of loading of CO2 for Mg2(dobdc) calculated at 40 °C using a dual-site Langmuir model. Reproduced with permission from ref 48. Copyright 2011 The Royal Society of Chemistry. shown in Figure 19, calculated using a dual-site Langmuir expression, shows a significant drop at a surface coverage that is consistent with the occupation approaching one CO2 molecule per Mg2+ site, which would be expected as a result of only weaker binding sites being vacant at this loading level. Note that the considerable difference in the isosteric heat of adsorption between the compounds is attributed to the higher ionic character of the Mg O bonds in Mg2(dobdc) compared with the Zn O bonds within Zn2(dobdc), leading to a higher partial positive charge on the Mg2+ metal centers, which consequently facilitates a greater degree of polarization on the adsorbed CO2 molecules.150 Importantly, the increased affinity for CO2 within Mg2(dobdc) leads to a higher CO2 uptake at low pressures and high temperatures (16.2 wt % at 0.15 bar and 323 K), which is crucial for post-combustion CO2 capture. More recently, Mg2(dobdc) and MOF-177 were evaluated for use in a temperature swing adsorption-based process by collecting single-component CO2 and N2 adsorption isotherms across a temperature range of interest for such a process (25 200 °C; see Figure 20).48 The working capacity of the materials was calculated as a function of the desorption temperature as the difference between the quantity of CO2 adsorbed at a partial pressure of 0.15 bar at the flue gas temperature (40 °C), and the corresponding value at 1 bar at the desorption temperature. For Mg2(dobdc), it was demonstrated that the working capacity for CO2 (shown in Figure 21) reaches 17.6 wt % for a 200 °C desorption temperature. The increase in working capacity with desorption temperature arises from the fact that the quantity of CO2 adsorbed is reduced at higher temperatures, leading to more complete evacuation of the pores during regeneration. A similar analysis performed on MOF-177 resulted in a negative CO2 working capacity across the entire temperature range. This is ascribed to the relatively limited CO2 adsorption capacity at low pressures due to the lack of strong binding sites within the pores, which affords a near-linear isotherm that steadily increases up to the desorption pressure. Even at the higher temperatures, a relatively high quantity of CO2 is retained at 1 bar, leading to a poor working capacity even at a desorption temperature of 200 °C. The results highlight the importance of selecting a structure containing a high density of strong binding sites within the pores, making Mg2(dobdc) a suitable candidate for further testing. While most metal organic frameworks possessing exposed metal cation sites are obtained by removing solvent molecules 751 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW CO2/N2 selectivity from 2.8 to 12.0, highlighting the success of this approach. While the chemical origin of the high performance of the Al(OH)(bpydc) 3 0.97Cu(BF4)2 compound has not yet been established, the increased affinity of the framework for CO2 is also demonstrated by an increase in the magnitude of the zerocoverage isosteric heat of adsorption from 23 to 30 kJ/mol, suggesting that a more suitable chemical environment for CO2 binding is facilitated by the insertion of Cu(BF4)2 into the channels of the framework. 3.3. Considerations for Application Many studies involving the synthesis and evaluation of metal organic frameworks for post-combustion CO2 capture applications focus exclusively on the CO2/N2 separation performance of the material, neglecting other important aspects of actual flue gas. While CO2 and N2 account for around 90% of the flue gas composition, understanding the effects of the gases present in the remaining 10% is critical to properly evaluating any material for use in a realistic CO2 capture process. Here, we discuss a number of studies that have aimed to study the performance of metal organic frameworks under more realistic conditions, including the use of humidified gas mixtures. 3.3.1. Stability to Water Vapor. While the primary challenge of CO2 capture is the separation of CO2 and N2, the detailed study of metal organic frameworks for post-combustion CO2 capture must take into account the fact that flue gas is saturated with H2O (5 7% by volume).11,31 While partial dehydration of the effluent may be possible, drying the flue gas completely prior to extracting CO2 is costly and most likely not feasible on such a large scale.31,367 As such, adsorbents used in CO2 capture from flue gas must be stable in the presence of at least some water vapor. In evaluating metal organic frameworks for applications in post-combustion CO2 capture processes, it is important to consider not only the stability of the framework to water vapor but also the effect of water vapor on the separation of CO2 from N2. With regard to water stability, the metal ligand bond is typically the weakest point of a metal organic framework, and hydrolysis can lead to the displacement of bound ligands and collapse of the framework structure (eq 1).155 This was first observed in the zinc-carboxylate based MOF-5, which is watersensitive and begins to lose crystallinity upon exposure to small amounts of humidity.115,385 387 Molecular dynamics simulations were used to show that the mechanism of hydrolysis for MOF-5 likely involves the direct attack of H2O at the tetrahedral Zn2+ centers to displace bound BDC2 ligands and destroy the framework structure.388 Figure 20. CO2 adsorption isotherms for Mg2(dobdc) collected over a temperature range of 20 200 °C. Reproduced with permission from ref 48. Copyright 2011 The Royal Society of Chemistry. Figure 21. Estimated working capacity as a function of desorption temperature (Td) for MOF-177, Mg2(dobdc), and zeolite NaX. The working capacity is calculated as the difference between the amount of CO2 adsorbed at 0.15 bar at a flue gas temperature of 40 °C and the amount of CO2 adsorbed at 1 bar at the desorption temperature. Reproduced with permission from ref 48. Copyright 2011 The Royal Society of Chemistry. from the metal centers after synthesis of the framework structure, one alternative strategy that has been pursued is the postsynthetic insertion of metal ions. For example, an aluminum-based metal organic framework Al(OH)(bpydc) (MOF-253, bpydc2 = 2,20 -bipyridine-5,50 -dicarboxylate) was synthesized, which exhibits the same network topology as Al(OH)(BDC) (MIL-53) and possesses open bipyridine sites along the one-dimensional pores.204 Owing to the low affinity of the bipyridine moieties toward the Al3+ ions used to construct the framework, these sites are vacant following synthesis, allowing metal ions to be selectively bound postsynthetically. The framework was loaded with a variety of transition metal salts, and the resulting products were screened for their CO2 adsorption properties. Of the compounds studied, the Cu2+-loaded material, Al(OH)(bpydc) 3 0.97Cu(BF4)2, exhibited the best performance, where the quantity of CO2 adsorbed at 1 bar at 298 K increased from 6.2 wt % in the bare framework to 11.7 wt % following metal insertion, despite a decrease in the BET surface area from 2160 to 705 m2/g. The increased uptake is accompanied by a dramatic increase in the ML þ H2 O f MðOHÞ þ LH ð9Þ Several attempts have been made at understanding the differences in relative water stability of metal organic frameworks. Highthroughput screening of frameworks exposed to different temperatures of steam was used to suggest that the strength of the metal oxygen bonds within the framework correlates with resistance to hydrolysis, with strong metal oxygen bonds imparting greater hydrothermal stability on the resulting framework.155 In a different study, representative metal organic frameworks were tested in different ratios of H2O to DMF, and the resulting differences in stability were attributed to the metal nodes within the framework.389 In agreement with previous work, it was found that the basic zinc acetate clusters characteristic of most zinc 752 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Figure 22. The general trend of increasing pKa for ligands built from carboxylic acids, tetrazoles, triazoles, and pyrazoles. The metal ligand bond is expected to be stronger as the pKa increases. carboxylate metal organic frameworks, such as the IRMOF series and MOF-177, are most susceptible to hydrolysis.390 The trinuclear chromium clusters found in many of the MIL series of frameworks are the most stable of the studied building units, while the copperpaddlewheel carboxylate clusters found in HKUST-1 exhibit intermediate stability. One of the primary efforts to increase the water stability of metal organic frameworks has been through the use of azolatebased linkers rather than the typical carboxylate linkers.200,391 397 The azolate linkers can bind metals with a similar geometry to carboxylate ligands, but their greater basicity typically leads to stronger M N bonds and greater thermal and chemical stability in the resulting framework. The relative M N bond strengths can be predicted based on the pKa values associated with the deprotonation of the free ligand. As such, stability typically decreases with the pKa: pyrazole (pKa = 14.4) linkers exhibit the greatest stability, imidazole (pKa = 10.0) and triazole (pKa = 9.3) are intermediate, and tetrazole (pKa = 4.6) linkers are the most labile (see Figure 22).398,399 Illustrating the remarkable stability of some of these frameworks, Ni3(BTP)2 (H3BTP = 1,3,5tris-1H-pyrazol-4-yl)benzene) is stable in boiling water at pH 2 14 for at least 14 days.397 An alternative strategy for increasing the metal ligand bond strength in metal organic frameworks is through the use of trior tetravalent metal cations.155 In general, frameworks containing Cr3+, Al3+, Fe3+, and Zr4+ cations display a high degree of stability in water. Specifically, MIL-53 (M(OH)(BDC), M = Cr3+, Fe3+, Al3+) is a flexible framework that expands or contracts based on the absence or presence of water.119,144,335 Here, the structural transition is reversible, and the overall framework scaffold remains intact upon repeated exposure to water. MIL100 and MIL-101 are rigid trivalent frameworks built from trinuclear metallic clusters that have shown a high stability in both boiling water and steam.400 404 Similarly, the zirconium(IV)-based metal organic framework UiO-66, which contains extremely robust Zr6O4(OH)4(CO2)12 cluster units (see Figure 23), has been shown to be very water-stable.117 Aside from increasing the metal ligand bond strength, other strategies have been successfully employed to synthesize metal organic frameworks with intrinsic functionalities designed to protect the material from hydrolysis. For instance, several Zn2+and Cu2+-based frameworks have been synthesized with highly hydrophobic surfaces that adsorb only very small amounts of water at low pressures.405 408 Additionally, several metal organic frameworks have been synthesized with water-repellant groups incorporated directly into the organic ligands as a means of protecting the metal core from water.409 411 In one promising study, a metal organic framework (Banasorb-22) isostructural to MOF-5 was synthesized, wherein each of the terephthalic acid linkers was modified with a trifluoromethoxy substituent.409 The resulting material demonstrated improved stability with regard to Figure 23. A portion of the crystal structure of the high-stability metal organic framework UiO-66.117 Yellow, gray, and red spheres represent Zr, C, and O atoms, respectively. H atoms are omitted for clarity. unmodified MOF-5. X-ray powder diffraction patterns indicated only minor changes to the framework structure after exposure to steam from boiling water for 1 week, while MOF-5 was shown to decompose into Zn3(OH)2(BDC)2 (MOF-69C) after just 2 h of exposure to steam. There have been several investigations of metal organic framework stability in liquid water, but few with regard to different levels of humidity and temperature. Note that due to the temperature and composition of the flue gas, stability to water vapor (rather than liquid water) is more relevant to CO2 capture applications.11 While several studies have reported water adsorption isotherms for various metal organic frameworks, most do not rigorously evaluate the stability of the framework upon repeated exposure to water vapor. In addition to the structural and chemical stability of metal organic frameworks in the presence of water vapor, evaluating any new compound for use in a post-combustion CO2 capture process also requires knowledge of the impact of water on the CO2/N2 separation performance. Indeed, small amounts of water can significantly affect the CO2 adsorption properties of many zeolites due to its strong adsorption on the highly hydrophilic surface, which prevents CO2 from interacting with many of the strong adsorption sites in the material.412 Initial efforts at understanding the effect of water on CO2/N2 separations in metal organic frameworks have focused on HKUST-1. It is well-known that this framework can reversibly bind water, and there have been extensive studies, both theoretical and experimental, on this topic.133,175,191,192,298,404,413,414 In its fully hydrated form, one water molecule is bound as a terminal ligand to each axial Cu2+ site. By heating of the material to 100 °C, these water molecules can be desorbed, generating the activated form of the framework as evidenced by a color change from turquoise (hydrated) to dark blue (dehydrated).133 The effect of water coordination on the CO2 adsorption properties of HKUST-1 was examined.191 In the fully hydrated form, less than 1 mmol/g of CO2 was adsorbed at 1 bar, compared with near 5 mmol/g for the dehydrated form. This is in agreement with a similar study that found that after exposure to 30% relative humidity, HKUST-1 experiences a decrease in CO2 uptake to 753 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW about 75% of its original value and a concurrent loss of some crystallinity.192 Surprisingly, when HKUST-1 was loaded with 1 equiv of water per every two Cu2+ sites, the CO2 adsorption increased to around 8 mmol/g, representing one of the highest CO2 uptakes of any metal organic framework at 1 bar (see Table 4).191 The increased CO2 capacity is attributed to electrostatic interactions between water bound to Cu2+ sites and the quadrupole moment of CO2. Unfortunately, a loading of one H2O molecule per two Cu2+ sites corresponds to such a low partial pressure of water that this form of HKUST-1 is unlikely to be realized in any realistic application. In a similar study, CO2 adsorption isotherms were measured at different water loadings for HKUST-1 and Ni2(dobdc).192 Both metal organic frameworks retained some capacity for CO2 at low water loadings but exhibited essentially no capacity for CO2 above 70% relative humidity (RH). Significantly, water adsorption resulted in a much faster decrease in CO2 uptake for zeolites 5A and NaX than for either metal organic framework. The effects of humidity on the performance of the M2(dobdc) (M = Zn, Ni, Co, and Mg) series of metal organic frameworks was studied in order to evaluate their CO2/N2 separation performance under conditions more relevant to post-combustion capture.326 A flow-through apparatus was used to measure the breakthrough CO2 capacity of each material in a 5:1 N2/CO2 mixture. The samples were then exposed to the same mixture at 70% RH, followed by regeneration at high temperatures. The breakthrough CO2 uptake capacity after regeneration was then measured (see Figure 24). The compound Mg2(dobdc), which has the highest reported capacity for CO2 at low pressures, performed the worst out of the series with a recovery of only 16% of its initial CO2 capacity after regeneration. Interestingly, this result stands in sharp contrast to previous reports that Mg2(dobdc) can be fully regenerated after exposure to water without any effect on its CO2 adsorption performance.190 The reason for these different results is not yet established, though the observations could be consistent with incomplete dehydration, which would dramatically lower the CO2 adsorption capacity. Nevertheless, Ni2(dobdc) and Co2(dobdc) performed far better with recoveries of 61% and 85% of their initial CO2 capacity, respectively. This is in agreement with a similar study that found Ni2(dobdc) was able to maintain its CO2 capacity after steam conditioning and long-term storage, while Mg2(dobdc) suffers a significant loss in capacity.415 The work serves to highlight the importance of looking at the effect of water on the CO2 adsorption properties of metal organic frameworks, because the uptake of CO2 as determined from single-component isotherms may not be the best indicator of material performance in a real-world post-combustion CO2 capture process. While strong adsorption sites, such as the exposed metal cations present in the M2(dobdc) series, can be poisoned by small amounts of water, some flexible metal organic frameworks have shown promising CO2 adsorption properties in the presence of water.266 In one report, water induced structural changes in MIL-53 that promoted a higher selectivity for CO2 over CH4.220 Similarly, the breakthrough CO2 adsorption of the flexible framework NH2-MIL-53(Al) in the presence of 5% water vapor was also studied.346 At low pressures, this material adopts a narrow-pore conformation that prevents the adsorption of certain gas molecules and improves its selectivity. Interestingly, CO2 is selectively retained by the framework even in the presence of water. Figure 24. Comparison of the flow-through CO2 capacities, as determined from breakthrough experiments using a 5:1 N2/CO2 mixture, for pristine M2(dobdc) and regenerated M2(dobdc) after exposure to 70% RH. Reproduced with permission from ref 326. Copyright 2011 American Chemical Society. All of the studies presented here have involved either the measurement of single-component CO2 adsorption isotherms after exposure to water or the measurement of breakthrough curves. While these experiments are crucial for the initial assessment of metal organic frameworks for post-combustion CO2 capture, multicomponent adsorption isotherms are of high priority for evaluating and understanding the performance under conditions likely to be encountered in an actual capture system.11 To the best of our knowledge, there are no direct measurements of multicomponent isotherms involving CO2 and H2O in metal organic frameworks to date. 3.3.2. Other Minor Components of Flue Gas. Although the gases CO2, N2, and H2O account for greater than 95% of the flue gas mixture, the other minor components (mostly O2, SOx, NOx, and CO) present cannot be ignored in assessing metal organic frameworks for post-combustion CO2 capture. While the exact amount of each species present in flue gas varies based on the type of coal burned and the specific configuration of a given power plant, a representative flue gas composition for a power plant burning low-sulfur eastern bituminous coal is reproduced in Table 1. Significantly, the effects of these trace gases on the CO2/N2 separation performance of metal organic frameworks remains unknown. Future research must also address this issue if metal organic frameworks are to be implemented in an actual post-combustion CO2 capture process. 4. PRE-COMBUSTION CAPTURE Pre-combustion CO2 capture is a process in which fuel is decarbonated prior to combustion, resulting in zero carbon dioxide production during the combustion step (see Figure 3). Here, coal is gasified, generally at high temperature and pressure, to produce synthesis gas or “syn gas”, which is a mixture of mostly H2, CO, CO2, and H2O (see Table 9).416 This gas mixture is then run through the water gas shift reaction to produce H2 and CO2 (“shifted syn gas”) at high pressure and slightly elevated temperature (5 40 bar and 40 °C, depending on the production plant).417 420 Pre-combustion CO2 capture, which refers to the separation of CO2 from H2 within this gas mixture, can then be performed to afford pure H2, which is subsequently combusted in a power plant to generate electricity. 754 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 9. Composition of Shifted Product of Coal Gasification421 component Table 10. Components of the Shifted Products of Steam Methane Reformation417 approximate percentage (vol %) component approximate percentage (vol %) CO2 25 35 H2 70 80 CO H2 0.5 0.7 30 50 CO2 15 25 CO 1 3 CH4 0 CH4 3 6 H2O 15 40 H2S 0.1 0.2 N2 0.3 2.3 Ar 0.04 responsible for the purification of millions of tons of hydrogen annually.418 This hydrogen is used for ammonia synthesis and large industrial-scale reactions such as the Fischer Tropsch synthesis of hydrocarbons. The ubiquitous nature of hydrogen purification via pressureswing adsorption motivates the use of this separation process in the context of pre-combustion processes in two ways. First, because the technology is already in use, pre-combustion CO2 capture would only require slight modifications or extensions to existing processes rather than a complete process overhaul. Second, from a practical point of view, the tremendous scale of H2 purification (50 million tons of H2 annually) suggests that even slight enhancements in the efficiency of this process would have significant benefits on the energy infrastructure as a whole. For example, if the energy cost of purification (7.3 MW 3 h per ton of H2) can be reduced by 10%, the energy savings would represent the equivalent of shutting down 18 average-sized coal-fired power plants.418,429 Thus, the development of next-generation materials for pre-combustion CO2 capture that exhibit improvements in energy efficiency over existing materials would also have a significant impact on the economic and environmental costs associated with H2 production worldwide. Zeolites and activated carbons (see section 1.4.2) remain the most well studied solid adsorbents for H2 purification of gas streams from steam-methane reformation.430,431 Zeolite 13X possesses a high selectivity for CO2, CO, and CH4 over H2 compared with BPL activated carbon,417 although the capacity of activated carbons is generally greater than that of zeolites. Considering that these are the state-of-the-art porous solid adsorbents used for hydrogen purification, we compare the performance of these materials with metal organic frameworks in the following sections to evaluate the potential of the latter to become the materials of choice for this application. 4.1.3. Metrics for Evaluating Adsorbents. In pressureswing adsorption processes, the performance of adsorbents is evaluated using the selectivity for CO2 over H2 (eq 6), sometimes in the presence of other gases, such as CO. Although the selectivity value is the most important factor for determining the purity of the respective gas phases following the separation process, an equally important metric for evaluating an adsorbent is the CO2 working capacity. This parameter is evaluated by taking the difference between the quantity adsorbed at flue gas pressure and the quantity adsorbed at the lower purge pressure. A high working capacity is particularly favorable in that less adsorbent is needed to form the PSA bed that performs the separation, which in turn decreases initial capital costs and, more importantly, lowers the long-term energy requirements for adsorbent regeneration. An increase in either the selectivity or the working capacity would decrease the cost of a pressure-swing adsorption-based CO2 capture system, and indeed, it has been reported that 4.1. Considerations for Pre-combustion CO2 Capture Solid adsorbents, membrane materials, and liquid absorbers are currently under consideration as potential candidates for use in pre-combustion CO2 capture.422 With regard to industrial applications, pre-combustion CO2 capture systems based upon CO2-absorbing solvents are the closest to being realized,423,424 and a number of power plants incorporating such systems are being constructed.425,426 Furthermore, the use of solid adsorbents in pressure-swing adsorption-based processes is currently under intense investigation. Here, the high-pressure gas mixture is transported through a packed bed of a porous adsorbent incorporated into a PSA process. The CO2 within the flue gas is selectively adsorbed along with other minor impurities of the mixture, such as CO, and following a series of concurrent and countercurrent pressurization and depressurization steps, the adsorbed gases can be removed, and the capture material can be regenerated for subsequent CO2 capture cycles.427 4.1.1. Advantages of Pre-combustion Capture. In comparison to post-combustion CO2 capture and oxy-fuel combustion-based processes, pre-combustion CO2 capture carries a number of advantages that may be of benefit for its rapid industrial implementation. Perhaps most significantly, the gases are produced at high pressure, and the partial pressure of CO2 in the mixture is high compared with post-combustion flue gas. As a result, regeneration of the loaded adsorbent can occur through a drop to atmospheric pressure, which is energetically favorable and is more practical compared with a temperature or vacuum swing-based process. Furthermore, the CO2/H2 separation is inherently easier to perform than CO2/N2 or O2/N2 separations, owing to the greater differences in the polarizability and quadrupole moment of the molecules.10 Thus, for purely physisorption-based separations, a greater selectivity for CO2 over H2 can be anticipated, which may allow next-generation precombustion CO2 capture materials to be more rapidly developed than new adsorbents for post-combustion CO2 capture or oxyfuel combustion. 4.1.2. Hydrogen Purification. Pressure-swing adsorption separation of shifted syn gas is a mature technology because of its wide use in hydrogen purification. Hydrogen used for industrial applications is generally synthesized from methane reformation followed by the water gas shift reaction, which is very similar to coal gasification. The same processes as for coal gasification are applied, although the composition of the final products differ in the partial pressures of the gases (see Table 10), owing to the differences in carbon content between methane and coal. While liquid solvents such as Selexol are often used,428 PSA columns employing solid adsorbent beds are also common and are 755 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW carbons and zeolites, and their tunable surface chemistry is anticipated to facilitate further optimization of the material properties. Here, optimization refers to creating adsorption characteristics that are ideal for the CO2/H2 separation with the aim of reducing the regeneration cost of the PSA adsorbent, while maintaining a high gas adsorption working capacity and selectivity for CO2 over H2.419 Despite the opportunity for the development of metal organic frameworks as pre-combustion CO2 capture adsorbents (and H2 purification adsorbents), relatively few reports have emerged in this regard. Although evaluation of the performance of candidate frameworks can be well-approximated via the collection of high-pressure, single-component CO2 and H2 isotherms at near-ambient temperature, there are only a small number of examples where such experiments have been performed. In cases where they have been reported, the isotherms have seldom been discussed and analyzed in the context of precombustion capture. Note that in this section, we consider CO2/ H2 separations only in the context of pressure-swing adsorptionbased processes in which the separation of the gases is achieved by a thermodynamic equilibrium that results from the bulk adsorptive properties of the material. As mentioned above, an alternative strategy for achieving the separation would be to make use of the difference in the kinetic diameters or diffusion properties of the two molecules in a kinetic-based separation using metal organic framework membranes, and a thorough discussion of CO2/H2 separations (and other gas separations relevant to CO2 capture) using such materials can be found in section 6. It should further be noted that, although the present discussion predominantly addresses CO2/H2 separations, CO2/ CO separation is also highly relevant to pre-combustion CO2 capture (see section 4.1.4) and has been investigated in metal organic frameworks such as ZIF-100247 and mixed-linker MOF-5 variants (multivariate “MTV” metal organic frameworks).219 The separation of CH4 and H2 is also highly relevant to hydrogen purification and has been investigated in metal organic frameworks in a number of studies.320,443 445 Table 3 provides high-pressure adsorption data for metal organic frameworks representing many structure types and surface chemistries, and in general the data demonstrate the tremendous CO2 adsorption capacity owing to the high surface areas and pore volumes found within the materials. In evaluating these materials for pre-combustion carbon dioxide capture, an initial step would be to calculate the maximum PSA working capacity of the potential adsorbent, which requires knowledge of its saturation capacity for CO2. While the adsorption capacity at the lower (purge) pressure is a crucial parameter for maximizing the working capacity, a high saturation capacity is a necessity since it serves as an upper limit of the working capacity. The selectivity of the material is also a crucial parameter when considering a material for pre-combustion CO2 capture. There can be a trade-off between the working capacity and the selectivity, since more selective materials will tend to have a steeper initial portion in the CO2 isotherms (P < 1 bar). This leads to a lower working capacity because the CO2 adsorbed at the lowest pressures will not be removed at the purge pressure. Thus, since the working capacity is highly sensitive to the lowerpressure region of the adsorption isotherm, a single high-pressure CO2 isotherm is not sufficient to evaluate a material for precombustion CO2 capture. For example, among the more promising metal organic frameworks listed in Table 3 are the MOF-74 optimization of both of these parameters is essential for developing promising candidates for PSA processes.432,433 4.1.4. Non-CO2 Impurities in CO2/H2 Streams. In both hydrogen purification and pre-combustion CO2 capture applications, the two main gas components are CO2 and H2. As can be seen in Tables 9 and 10, both product gases also contain CO, and in the case of steam-methane reformation, CH4 is also present. Thus, discovery of materials that selectively capture CO and CH4 in the presence of H2 is also of high importance, since the impurity gases adsorb onto surfaces less strongly than CO2 and conseqently break through the end of the column with H2.434 Both of these impurities must be removed owing to the implications of CH4 as a greenhouse gas435 and the high toxicity of CO. Other trace impurities, such as H2O and H2S, may also influence the performance of the material over time, and at the very least, the materials should be robust toward these components. The stability of metal organic frameworks toward H2O was considered in the context of post-combustion CO2 capture in section 3.3.1. There is relatively little known regarding the effect of H2S on metal organic frameworks, although initial studies have demonstrated strong (irreversible) adsorption within materials with exposed metal cation sites, such as MIL-100.436,437 4.1.5. Metal Organic Framework-Containing Membranes for pre-combustion CO2 Capture. Membrane separation is an exceptionally promising strategy for precombustion CO2 capture.438 This is primarily because the high pressure of a pre-combustion gas mixture is an excellent driving force for membrane separation of CO2 and H2. The same properties that make metal organic frameworks promising precombustion adsorbents are preserved in a membrane separation scenario. In a MOF-5 membrane, the CO2/H2 selectivity is theoretically higher than the CO2/N2 selectivity.313 The high capacities for CO2 in bulk metal organic frameworks are not necessarily lowered when these materials are incorporated into membranes, since, for example, mixed-matrix membranes of MIL-53 have been shown to adsorb as much as 6 mol/kg of CO2 at 25 bar.439 Diffusion is a property of gases that is pertinent to membrane separations that is frequently not discussed for bulk nanoporous adsorbents. This important characteristic has been thoroughly studied and can be harnessed in CO2/H2 membrane separations in metal organic frameworks.440 In MOF-5, the self-diffusivity of H2 is much higher than that of CO2, a requisite property for a H2-selective membrane.441 H2 diffusion in MIL-47(V) and MIL-53(Cr) has been shown to be 2 orders of magnitude higher than the highest values recorded within a zeolite, which is a potential advantage in incorporating metal organic framework-containing membranes into pre-combustion CO2 capture systems. Additionally, CO2 diffusion is slower than H2 in MIL-47 just as in MOF-5,442 preserving the selectivity of the separation materials. In section 6, we will discuss membrane applications of metal organic frameworks in more detail. 4.2. Metal Organic Frameworks as Adsorbents Metal organic frameworks have recently been investigated as potential next-generation adsorbents for pressure-swing adsorption-based separation of CO2 from H2. Their high surface areas afford enhanced gas adsorption capacities compared with porous solids conventionally employed in multilayer beds within current PSA systems, namely, activated 756 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Figure 26. CO2/H2 selectivities for an 80:20 H2/CO2 mixture at 313 K as calculated using IAST methods. Reproduced with permission from ref 87. Copyright 2011 American Chemical Society. Co-BDP) were found to display low CO2/H2 selectivities in the range of 5 10. In terms of working capacity for the same gas composition, all of the metal organic frameworks exhibited higher volumetric and gravimetric working capacities compared with zeolite 13X and activated carbon JX101 (see Figure 27). The high CO2 saturation capacities of the metal organic frameworks, especially coupled with strong polar interactions (such as in the case of Mg2(dobdc) and Cu-BTTri) lead to their high working capacities. Note that the materials with large pores (MOF-177 and Be-BTB) have a high gravimetric capacity but a low volumetric capacity, highlighting the importance of structural considerations when selecting a metal organic framework for pre-combustion CO2 capture. The results of this study suggest that metal organic frameworks with open metal sites or other sources of localized charge are particularly promising as pre-combustion CO2 capture adsorbents. They offer high CO2/H2 selectivities as well as high working capacities. Due to the delicate balance between strong CO2 binding and the need for low CO2 adsorption at low pressures (in order to increase working capacity), screening a wide variety of this class of metal organic frameworks is of high importance for identifying the best materials for this application. 4.2.2. Computational Studies. While the performance of metal organic frameworks can be probed by collecting highpressure CO2 and H2 isotherms, an alternative option for screening the vast number of materials would be to simulate the adsorption of a pre-combustion gas mixture to predict the separation performance. This can be performed by using only the crystal structure of a porous material, making it a very powerful tool. While such a methodology would allow a large number of structures to be rapidly screened, it should be noted that synthetic limitations (such as obtaining a perfectly desolvated sample) make theoretical results a best-case scenario that require experimental substantiation in the case of the most promising materials. In a similar situation to the experimental studies discussed above, simulation-based investigations of CO2/H2 separations are also limited to a relatively small number of reports. One of the earliest reports involved an indium-based metal organic framework, [In3O(abtc)1.5(H2O)3](NO3) 3 3H2O, consisting of 3,30 , 5,50 -azobenzenetetracarboxylate (abtc) organic bridging units and charge-balancing nitrate anions in the pores.449 The predicted selectivity for CO2 over H2 in a 15:85 CO2/H2 mixture introduced at 298 K increases from approximately 300 to 600 Figure 25. Single-component CO2 (green triangles) and H2 (blue circles) adsorption isotherms for the metal organic frameworks MOF177, Co(BDP), Cu-BTTri, and Mg2(dobdc) at 313 K. Reproduced with permission from ref 87. Copyright 2011 American Chemical Society. structure types (Mg- and Ni-MOF-74), which possess exposed metal cation sites (see section 3.2.3). These materials do not have the highest CO2 saturation capacity and also have a relatively steep CO2 adsorption isotherm, but as will be discussed in section 4.2.1, the balance between selectivity and working capacity renders them extremely promising for pre-combustion CO2 capture. In contrast, MOF-177 has a high CO2 capacity and a shallow initial rise in its CO2 adsorption isotherm. However, its internal surface imparts little CO2/H2 selectivity and after further investigation, MOF-177 is shown to be a poor pre-combustion CO2 capture material. These two examples serve to stress that while metal organic frameworks adsorb large amounts of CO2 at high pressures, they need to be carefully evaluated in order to elucidate the most promising candidates for pre-combustion CO2 capture. 4.2.1. Investigations Based on Single-Component Isotherms. To date, very few experimental studies explicitly investigating metal organic frameworks or pre-combustion CO2 capture have been reported. In one recent example, high-pressure CO2 and H2 adsorption isotherms were measured at 40 °C for five rerpresentative metal organic frameworks, and the selectivities and working capacities were calculated using IAST.87 Two rigid, high-surface area frameworks (MOF-177 and BeBTB),446,447 one flexible framework (Co(BDP)),448 and two materials bearing exposed metal cation sites (Mg2(dobdc) and Cu-BTTri)150,200 were studied. Figure 25 shows the isotherms for four of the five frameworks (the isotherms for Be-BTB were similar to those obtained for MOF-177). In terms of adsorption selectivity, Mg2(dobdc) was the best out of the materials studied, including two activated carbons and zeolite 13X (see Figure 26). These selectivities were calculated by applying IAST (section 2.3.2) to the single-component isotherms and assuming an 80:20 H2/CO2 mixture. All porous materials with localized charges in the pores (Mg2(dobdc) and Cu-BTTri) and zeolite 13X demonstrated selectivities between 100 and 1000, as did activated carbons with strong van der Waals interactions with CO2 due to their small pores. The metal organic frameworks having largely aromatic internal surfaces without significant surface charges (MOF-177, Be-BTB, and 757 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Figure 27. Gravimetric (left) and volumetric (right) CO2 working capacities calculated using IAST for an 80:20 H2/CO2 mixture at 313 K, assuming a purge pressure of 1 bar. Reproduced with permission from ref 87. Copyright 2011 American Chemical Society. between 0 and 5 bar and then gradually decreases to 450 up to 30 bar. While the authors did not report the CO2 working capacity for this material, the high-pressure CO2 adsorption of a 15:85 CO2/H2 mixture was reported and from this isotherm a maximum working capacity of 7 mmol/g can be approximated. This is similar to the highest performing metal organic frameworks discussed in section 4.2.1, and therefore, if the nitrate anions are accessible to gas molecules in the experimentally prepared material, this compound would be very promising for pre-combustion CO2 capture. Quaternary and quinary mixtures were also simulated in this study. The quaternary mixture was composed of 15:75:5:5 CO2/ H2/CO/CH4 and the quinary mixture was 15:75:5:5:0.1 CO2/ H2/CO/CH4/H2O. These were included to better model precombustion carbon capture mixtures as discussed in section 4.1. Both selectivity curves maintain the same shape as the binary mixture; however, the maximum selectivity drops by approximately 20 for the quaternary and another 30 for the quinary mixture. The selectivity continues to decline from 5 to 30 bar, and as such, the quaternary mixture drops to approximately 375 and the quinary selectivity to 350. In the context of pre-combustion CO2 capture, this work suggests that a modest amount of water in the gas mixture could significantly decrease the selectivity for CO2 due to selective binding of H2O over CO2 due to its stronger polarizability. Interestingly, between 0 and 2 bar the highest selectivity was observed for the quinary mixture,which was attributed to bound water promoting the adsorption of CO2. The same four- and five-component mixtures were employed in evaluating a second cationic metal organic framework with charge-balancing nitrate anions.450 The copper-based material, Cu6O(TZI)3(H2O)9(NO3) 3 15H2O (rht-MOF), was shown to be much less selective for CO2 in both the quaternary and quinary mixtures, and the initial selectivity of ca. 42, drops to 40 at 1 bar, increases to 60 at 40 bar, and then approximately plateaus until 50 bar. The authors attributed this difference to the large pore volume of rht-MOF. A tetracarboxylate-linked indium-based metal organic framework was also investigated for CO2/H2 separation.451 This anionic framework with formula Li0.5(H3O)0.5[In(C16H6O8)] contains charge-balancing Li+ cations in the pores, which are shown theoretically to be the primary binding site for CO2. The selectivity for CO2 in a 15:85 CO2/H2 mixture decreases from 1100 to 600 between 0.01 and 1 bar, increases slightly until 10 bar, and then decreases back down to 500 at 50 bar (see Figure 28). These three regions are attributed to strong binding sites becoming unavailable with increasing CO2 loading, followed by cooperative Figure 28. Selectivities calculated from simulations for a 15:85 CO2/H2 mixture at 298 K in Li0.5(H3O)0.5[In(C16H6O8)] shown with (solid line) and without (dashed line) inclusion of electrostatic interactions. Reproduced with permission from ref 451. Copyright 2011 American Chemical Society. CO2 CO2 interactions becoming predominant, and last the entropic favorability of H2 binding at high pressures. This entropic effect has been observed previously for other porous materials.452 Simulations for CO2/H2 separations were further reported within an indium-based metal organic framework, In48(C5N2O4H2)96Na48(C2H5OH)96 (H2O)192 (rho-ZMOF), formed from 4, 5-imidazoledicarboxylic acid as the linker.453 This framework is anionic and contains charge-balancing Na+ cations in the pores, which bind CO2 strongly and preferentially. Only at high pressures when the cations are completely solvated by CO2 does the CO2 begin to bind at other surfaces. This strong Na+ CO2 interaction leads to an initial selectivity of 200 at 10 bar, which remains steady until 30 bar for a 15:85 CO2/H2 mixture. A maximum working capacity of 4 mmol/g can be estimated from the high-pressure CO2 adsorption of a 15:85 CO2/H2 mixture. This is approximately half that of other metal organic frameworks with exposed metal sites (see Section 4.2.1). Interestingly, the Na+ cations shift locations slightly with CO2 adsorption. The effect of replacing the protons on 1,4-benzenedicarboxylate (BDC) in MOF-5 (see Section 1.4.1, Figure 6) with O Li+ groups was studied computationally in a recent report.350 Upon exposure to a 20:80 CO2/H2 mixture, this structure reaches predicted CO2/H2 selectivities of 10 000 at infinite dilution, compared with approximately 10 for unmodified MOF-5. This was determined to be solely from electrostatic interactions by turning these forces off and examining the changes in selectivity. 758 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW The selectivity for CO2 in a 50:50 CO2/H2 mixture in HKUST-1 (see Figure 7) and MOF-5 were compared via simulations.316 In MOF-5, the authors found that the selectivity slowly increases from less than 10 to approximately 30. HKUST-1 demonstrates a selectivity curve similar to rht-MOF, with an initial decrease from 100 to 80 by 1 bar, an increase to 150 by 15 bar, and then a decrease to 100 at 50 bar. The last decrease is ascribed to the favorable packing of the small hydrogen molecules when gas molecules are filling the pores, as opposed to physisorption onto the surface. This effect is not seen in MOF-5, because at 40 bar the pores are not fully occupied. The authors also compared these selectivities to those generated using IAST and found that for MOF-5 the prediction is reasonably accurate but in HKUST-1 IAST underestimates the selectivity with a drop from 100 at 0 bar to 70 at 50 bar. An additional insight was the composition-dependence of selectivity, which was shown to be minimal as the selectivity did not change more than 10% upon increasing the amount of CO2 in the mixture from 10% to 90%. A compound with a neutral framework structure, Co2(ade)2(CO2CH3)2 3 2DMF 3 0.5H2O (bio-MOF-11), was studied theoretically for pure-component CO2 and H2 adsorption and a 15:85 CO2/H2 mixture.315 Radial distribution functions demonstrated that CO2 binds at Lewis basic sites of the adenine linker and not near the framework cobalt atoms. The selectivity of the gas mixture increases from 270 to 370 between 0 and 5 bar, then decreases to 220 at 30 bar. The addition of 0.1 mol % H2O to the mixture had a negligible effect on the selectivity, in contrast to the results of the cationic materials discussed above. Some researchers have focused on comparing groups of porous solids to extract meaningful trends in CO2/H2 separations. Simulated breakthrough curves (see section 3.2.3) of CH4/H2, CO2/H2, and CO2/CH4/H2 mixtures of five metal organic frameworks (as discussed in section 4.2.1), as well as zeolite 13X, were published recently.434 A packed bed of metal organic framework was simulated and both the composition of the gas mixture exiting the bed as a function of time and the amount of CO2 adsorbed were evaluated. The metric τbreak, the time at which a specified fraction of impurity (either CO2 or CH4) is present in the gas leaving the bed, was used to characterize the adsorbents. In terms of τbreak and the amount of CO2 adsorbed at τbreak, Mg2(dobdc) was determined to be the best material of those studied (see Figure 29). Note that this method provides an important means of ranking materials via a single metric, which takes into account both selectivity and working capacity. The effect of catenation (interpenetration) on CO2/H2 selectivity was examined computationally, and it was found that catenated IRMOFs are much more selective than the noncatenated materials.454 Looking at five different mixture compositions (5:95, 30:70, 50:50, 70:30, and 95:5 CO2/H2), the catenated metal organic frameworks all displayed selectivites between approximately 40 and 110, whereas the three noncatenated materials were much lower (below 20). The shapes of the selectivity curves differed for the different compositions, with the three steps discussed above for Li0.5(H3O)0.5[In(abtc)] becoming more apparent with more CO2 in the mixture. This substantiates the attribution of CO2 3 3 3 CO2 interactions to the increase in selectivity after an initial decrease, as this increase is not apparent in the 5:95 CO2/H2 mixture. Furthermore, turning off electrostatic interactions between CO2 molecules only resulted in the loss of this feature in the selectivity trace. The adsorption sites of the gas molecues of the mixture gas Figure 29. Amount of CO2 adsorbed as a function of breakthrough time (τbreak), where the breakthrough concentration of CO2 is 0.05% and the conditions are 313 K, 48 bar adsorption, and 12 bar desorption pressures. Reproduced with permission from ref 434. Copyright 2011 American Chemical Society. demonstrated that selective CO2 adsorption happens in the small channels of the catenated frameworks. A study in which the CO2/H2 selectivity for two ZIF materials, ZIF-3 and ZIF-10, was compared with those of IRMOF-1, -8, -10, and -14 has been recently reported.308 In a 10:90 mixture of CO2/H2, ZIF-3 exhibited the three-step adsorption selectivity behavior discussed above, with a selectivity factor between 100 and 200, while ZIF-10 increased linearly in selectivity from 25 to 60 between 0 and 50 bar. All of the IRMOF selectivities were approximately constant with pressure, varying between 7 and 12. The difference in selectivity between the two classes of frameworks was attributed to the smaller pore size in ZIFs and confinement effects, which is a result of the higher degree of overlap of the van der Waals forces associated with opposing wall surfaces. In a separate study, six metal organic frameworks were compared with six covalent-organic frameworks (COFs). These are porous crystalline solids that consist of only covalent bonds and are typically assembled from the co-condensation of boronic acids.314 The six COFs studied are very similar with respect to their properties since they are composed of the aromatic units connected by boron oxide rings. Five of the six metal organic frameworks studied are from the IRMOF series and the sixth was HKUST-1. Two materials, HKUST-1 and COF-6, were found to be much more selective than the other materials. The selectivity in COF-6 was attributed to the small pore size, and the concomitant confinement effects. Both materials had large differences in the heats of adsorption of CO2 compared with H2, and HKUST-1 exhibited a significantly higher electrostatic contribution to the selectivity. A compilation of previously published data obtained from configurational-bias Monte Carlo (CBMC) simulations was recently reported.317 The CO2/H2 selectivity for a 15:85 CO2/ H2 mixture was compared for zeolites NaX, NaY, AFX, CHA, and FAU-Si in addition to the metal organic frameworks rhoZMOF, Mg2(dobdc), Zn2(dobdc), HKUST-1, ZIF-8, and MOF-177. The most selective was zeolite NaX followed by NaY, AFX, rho-ZMOF, and Mg2(dobdc). The working capacities assuming a 1 bar purge pressure were also plotted for these materials and the same gas mixture, and Mg2(dobdc), Zn2(dobdc), HKUST-1, ZIF-8, and zeolite AFX were the top five 759 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW in terms of working capacity, with Mg2(dobdc) being consistently significantly higher than all of the other materials (see Figure 30a). In addition, selectivity as a function of working capacity was also plotted at a pressure of 10 bar to demonstrate which materials maintained a high working capacity and were also selective (Figure 30b). Here, Mg2(dobdc) stands out as clearly the most attractive metal organic framework candidate for pre-combustion CO2 separation. Taking all of the simulation work published to date relating to metal organic frameworks for pre-combustion CO2 capture into consideration, Mg2(dobdc) is expected to provide the best overall performance out of the materials studied so far. Importantly, this is consistent with the conclusions from the few experiments performed thus far (see section 4.2.1). Strategies for improving upon the performance of Mg2(dobdc) are urgently required, although the synthesis of new metal organic frameworks with pore surfaces decorated with a high concentration of exposed high-valent (trivalent or tetravalent) cations would be one approach for achieving higher capacities and selectivities. As discussed in section 4.1.4, examining the effects of impurity components (H2O, CO, H2S, and CH4) in the precombustion gas stream on the stability and separation performance in Mg2(dobdc) as well as other metal organic frameworks is an area in which a greater understanding is urgently needed. 5. OXY-FUEL COMBUSTION Oxy-fuel combustion refers to the ignition of pulverized coal or other carbon-based fuels in a nearly pure O2 environment and represents a relatively new process for mitigating CO2 emissions compared with pre-combustion and post-combustion CO2 capture. The significant advantages of this process stem from the fact that the flue gas (following removal of particulates, water, and trace impurity gases) is almost entirely CO2, which greatly simplifies the capture step, and that most existing power plants could be readily retrofitted with an oxy-fuel combustion system. Although there are no full-scale plants currently using oxy-fuel combustion, theoretical studies along with laboratory and pilotscale studies have provided an understanding of important design parameters and operational issues. A schematic of a typical oxy-fuel combustion process is presented in Figure 3. Here, in a conventional set up, O2 (purity >95%) is fed into the plant from a cryogenic separation unit, which separates O2 from the other components of dry air by a distillation process. The O2 inlet gas is diluted with CO2 from the flue stream to a partial pressure of 0.21 bar in order to control the temperature of fuel combustion and to reduce the formation of NOx impurities that frequently form when coal is burned in an O2-enriched atmosphere.455 The exhaust gas, which is essentially pure CO2, can then be directly subjected to sequestration using the techniques discussed in detail in section 1.3. Indeed, in addition to CO2 (55 65 wt %), the other major component of the gas stream is water vapor (25 35 wt %), which is easily condensed and removed.456 In fact, CO2 capture rates higher than 95% have been achieved by this method, a level not currently possible with precombustion and post-combustion separations.457 A further advantage related to combustion in an O2/CO2 mixture lies in the fact that compared with a process utilizing air, which is rich in N2, the formation of NOx is largely inhibited, allowing for a Figure 30. (a) Working capacities (“delta loadings”) of selected metal organic frameworks and zeolites, as calculated for a 15:85 CO2/H2 mixture at 300 K with a purge pressure of 1 bar. (b) The CO2/H2 adsorption selectivity as a function of working capacity (“delta loadings”) of selected metal organic frameworks and zeolites. The selectivities and working capacities are calculated for a 15/85 CO2/H2 mixture at at 300 K, with a purge pressure of 1 bar. Reproduced with permission from ref 317. Copyright 2011 The Royal Society of Chemistry. smaller, cheaper NOx removal step than required in current power plants. A significant challenge for the implementation of oxy-fuel combustion methods is in the large-scale generation of pure O2 from air. This separation is currently carried out on a scale of over 100 Mt/year,458 but the large energy requirement for this process creates an urgent need for alternative separation methods if oxy-fuel combustion is to be widely used in mitigating carbon emissions. Microporous solids that selectively adsorb O2 from the air could potentially significantly reduce this energy cost. Indeed, zeolites have been used in this separation on an industrial scale and in portable medical devices,459 although the separation performance and energy efficiency is considered to be insufficient for use in oxy-fuel combustion applications. In this regard, metal organic frameworks hold tremendous promise for delivering high-performance materials that are specifically optimized for the removal of O2 from the air. As will be discussed, this separation is essentially an O2/N2 separation, for which metal organic frameworks exhibiting high selectivities, and O2 adsorption capacities have been recently reported. 760 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW 5.1. Metal Organic Frameworks for O2/N2 Separation As discussed in previous sections, microporous metal organic frameworks are typically selective for the adsorption of CO2 over N2 or CO2 over H2, based on the large differences in the quadrupole moments and polarizabilities between the molecules. However, owing to the very similar physical properties of O2 and N2, their separation using a purely physisorptive adsorption mechanism is expected to yield very limited selectivity factors. In fact, because of the higher quadrupole moment and polarizability of N2 compared with O2 (see Table 2), many of the metal organic frameworks for which O2 isotherms have been measured at room temperature are slightly selective for the adsorption of N2 over O2. For example, O2 and N2 adsorption isotherms recorded at 298 K for MOF-177 and UMCM-1, which are both high-surface area frameworks with BET surface areas of greater than 4000 m2/g, yield a selectivity factor (calculated as the number of moles of O2 adsorbed at 0.21 bar divided by the number of moles of N2 adsorbed at 0.79 bar) of less than 1.169,228,298,390 Furthermore, although small-pore zeolites have been employed for O2/N2 separations based on the differing kinetic diameters of these molecules, it is unlikely that metal organic frameworks could offer the significant improvements needed for an efficient kinetic-based O2/N2 separation process. Instead, an effective separation of O2 from N2 within metal organic frameworks is expected to be achieved through harnessing the differences in chemical properties of the molecules, and this has indeed been the focus of recent work in which high adsorption selectivities have been reported. The high propensity for O2 to accept electrons from redoxactive metal centers is one feature that can be exploited to perform the separation of O2 from N2 within metal organic frameworks. Indeed, such interactions are crucial in nature for biological functions such as O2 transport, and a large volume of biomimetic molecular chemistry has emerged through coordination complexes containing metal ions such as Fe2+ or Cu+.460 468 Thus, if the surfaces of metal organic frameworks can be decorated with electron-rich, redox-active metals with open coordination sites, these sites may engage in a reversible electron transfer to O2, but not N2. Metal organic frameworks containing such accessible redoxactive metal centers have recently emerged,140,254,469 471 and these display considerable promise as next-generation O2 capture materials. Note that metal organic frameworks are expected to possess several advantages over zeolites, since their larger pores are expected to facilitate higher gas permeability, and the adjustable character of the pore surfaces are anticipated to allow the enthalpy of O2 adsorption to be finely tuned for specific working conditions. Additionally, since the exposed metal sites are immobilized and separated from each other on the pore surface by virtue of the framework structure, two metal sites cannot combine to form O2-bridged species, a reactivity pathway that is known to diminish the performance of molecular complexes investigated for O2 binding.472 474 Recently, a metal organic framework bearing exposed Cr2+ adsorption sites, Cr3(BTC)2, a framework isostructural to HKUST-1 was studied for selective binding of O2 over N2.134 The Cr2+ sites are available for binding guest molecules and exhibit a tremendous affinity for O2, leading to an O2 adsorption of 11 wt % at 298 K and a pressure of just 2 mbar (see Figure 31). At the same temperature, the quantity of N2 adsorbed is just 0.58 wt % at a pressure of 1 bar. The steep initial rise in the O2 adsorption isotherm is indicative of a strong (chemical) interaction between Figure 31. Uptake of O2 (red symbols) and N2 (blue squares) by Cr3(BTC)2 at 298 K. The compound saturates with O2 at approximately 2 mbar but shows little affinity for N2. Upon evacuation, the O2 isotherm for a second cycle reveals a slightly reduced capacity. Reproduced with permission from ref 134. Copyright 2010 American Chemical Society. the Cr2+ sites and O2 molecules, and the saturation uptake corresponds to the adsorption of dioxygen at approximately 80% of available Cr2+ sites within the framework. Indeed, UV vis nIR spectroscopy and X-ray absorption spectra confirmed the partial electron transfer from Cr2+ to the bound O2 molecule to give roughly a Cr3+ superoxide adduct. The charge transfer gives rise to a significantly stronger interaction compared with the physisorptive adsorption of N2 molecules, resulting in an O2/N2 selectivity factor of 19.3. This selectivity is significantly higher than that observed for most previously investigated porous solid materials,475 such as a carborane-based porous Co2+-based framework, which displays a selectivity factor of 6.5 at low pressure.254 Importantly, Cr3(BTC)2 displays reversible O2 uptake, and heating of the framework to just 50 °C liberates the majority of the bound O 2 . However, following 14 adsorption/desorption cycles, approximately 35% of the O 2 adsorption capacity of the material is lost, presumably as a result of framework decomposition owing to the highly exothermic nature of the chromium O2 interaction. More recently, selective O2 adsorption has been studied in Fe2(dobdc), a metal organic framework featuring one-dimensional hexagonal channels lined with five-coordinate Fe2+ centers (see Figure 8).476 Gas adsorption isotherms measured at 298 K indicate that this material binds O2 at a capacity of 10.4 wt % at 1.0 bar. The N2 uptake under these conditions is considerably lower, reaching just 1.3 wt % at 1.0 bar. Although this material exhibits a selectivity factor of 6.6, the adsorption of O2 at 298 K was irreversible. Remarkably, at 211 K, O2 adsorption was nearly doubled to 18.2 wt % and fully reversible. Under these conditions, the material showed negligible loss in adsorption capacity after 13 cycles. Initial isosteric heats of N2 ( 35 kJ/mol) and O2 ( 41 kJ/mol) adsorption in Fe2(dobdc), calculated from dual-site Langmuir Freundlich fits to isotherms measured at 201, 211, 215, and 226 K, reflect the higher propensity for O2 to accept charge from Fe2+. Accordingly, Fe2(dobdc) exhibits high O2/N2 selectivity at these temperatures ranging from 4.4 to 11, reaching a maximum of 11.4 at 201 K and about 0.4 bar. Application of IAST to simulate breakthough curves indicate Fe2(dobdc) is a promising material for the separation of O2 from air at temperatures well above those currently used for cryogenic distillations. 761 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW M€ossbauer spectroscopy, in conjunction with neutron powder diffraction and infrared spectroscopy, was employed to elucidate the mechanism of O2 adsorption in Fe2(dobdc). At low temperatures, Fe2(dobdc) was found to adsorb O2 reversibly at a capacity corresponding to the adsorption of one O2 molecule per Fe center. At higher temperatures O2 adsorption becomes irreversible and decreases to one O2 molecule per two Fe centers. M€ossbauer spectroscopy indicated that in the absence of O2, the spectrum of Fe2(dobdc) exhibits a simple quadrupole doublet, with isomer shifts and quadrupole splitting indicative of high-spin Fe2+ in a square pyramidal coordination environment. Upon dosing the material with O2 at 94 K nearly all of the iron centers display a substantially reduced isomer shift, which lies between typical values expected for high-spin Fe2+ and high-spin Fe3+, suggesting partial charge transfer to the adsorbed O2. Warming the material above 222 K in the presence of O2 results in spectra with isomer shifts and quadrupole splittings typical of high-spin Fe3+. The combination of data from O2 adsorption experiments and M€ossbauer spectroscopy suggests that O2 binds to the exposed iron sites in Fe2(dobdc) by two different coordination modes depending on the temperature of adsorption. At low temperatures, partial charge transfer from iron to O2 results in a material with weakly bound, partially reduced O2 at every Fe center. At higher temperatures, the initial electron transfer step is followed by a second electron transfer from an adjacent Fe center. In this scenario, all of the iron centers are oxidized to Fe3+, half being irreversibly coordinated by an O22 anion, while the other half remain five-coordinate. Infrared spectroscopy and neutron powder diffraction were utilized to confirm the proposed mechanism of O2 binding in Fe2(dobdc). Upon dosing the sample with O2 at 100 K, a number of new bands were apparent in the difference IR spectra. Most notable was the component present at 1129 cm 1, which, in conjunction with the first overtone of this stretching mode appearing at 2238 cm 1, is indicative of partially reduced (near superoxo) O2 species coordinated to the Fe centers. Additionally, upon warming the sample to room temperature, the superoxo band at 1129 cm 1 disappears while a new component at 790 cm 1 (indicative of metal-bound O22 ) appears. These assignments are consistent with the model developed from O2 adsorption experiments and M€ossbauer spectroscopy. Direct structural evidence for temperature-dependent binding of O2 to the iron centers in Fe2(dobdc) was provided by neutron powder diffraction. Refinement of powder diffraction data collected on a sample at 4 K that was dosed with 2 equiv of O2 per iron at 100 K indicate three different O2 adsorption sites (Figure 32). The first site to be populated by O2 is the open Febased binding site. Interestingly, the O2 molecule binds in a symmetric side-on coordination geometry, with an Fe O distance of approximately 2.10(1) Å. The refined O O separation distance of 1.25(1) Å is approximately halfway between the internuclear distance observed for typical superoxide (1.28 Å) and free O2 (1.207 Å).477 The second and third O2 adsorption sites occur within the pores of the material at occupancies of 0.857(9) and 0.194(8) Å, respectively. Refinement of data collected at 4 K on a sample that had been dosed with O2 at room temperature confirmed coordination of a peroxide species to approximately half of the iron centers in the framework. Specifically, O2 was found to coordinate to iron in an asymmetric side-on mode with a refined occupancy of 0.46(2) and a substantially elongated O O distance of 1.6(1) Å. Figure 32. A portion of the structure of Fe2(O2)2(dobdc) 3 O2 as viewed down the (001) direction. Orange, gray, and red spheres represent Fe, C, and O atoms, respectively; H atoms are omitted for clarity.477 The O2 adsorption characteristics of Fe2(dobdc) and Cr3(BTC)2 highlight the importance of exposed redox-active metal centers for the development of new metal organic frameworks that exhibit selective O2 adsorption. A simple strategy for preparing new metal organic frameworks with potential utility for O2/N2 separations is exemplified by the synthesis of these materials. Although Cu3(BTC)2478 or Mo3(BTC)2137 do not selectively adsorb O2 over N2, replacement of the metal cations in the structure of these materials with Cr2+ effectively resulted in a material with unprecedented O2/N2 selectivity. This approach of preparing frameworks of known structure types with redox-active metals could potentially be expanded to other families of materials. In addition to preparing new metal organic frameworks for O2/N2 separations, modifications can be made to existing frameworks to optimize the O2 adsorption selectivity. The compound Fe2(dobdc), for example, exhibits reversible O2 adsorption up to 226 K, above which electron transfer reactions result in irreversible O2 adsorption. Modification of the bridging ligand by incorporating electron-donating or -withdrawing groups would presumably alter the electronics of the material and could result in a framework that demonstrates reversible room-temperature O2 adsorption. Additionally, a number of metal organic frameworks have been shown to be amenable to the postsynthetic insertion of metal cations.204,479 483 Incorporation of redoxactive metals into these frameworks may lead to materials with selective/reversible O2 binding properties. As compared with CO2 capture and storage, O2/N2 separations with metal organic frameworks are still relatively new. However, rapid progress has been made in this area, and further work is expected to lead to a large class of O2 separation materials that operate with high capacity at room temperature. 6. METAL ORGANIC FRAMEWORK-CONTAINING MEMBRANES Gas separation using membranes is a kinetics-based process that relies on differences in the diffusion rates of gas molecules within the membrane materials. In addition to the chemical interactions between gas molecules and the surfaces of the 762 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW pure film, or as one component within a mixed-matrix membrane. Although our focus will be predominantly on those separations that would be of high importance for reducing CO2 emissions from power plants, a number of other CO2 separations, such as CO2/CH4 separations, that illustrate the important considerations for the development of next-generation membrane materials for CO2 capture, will also be discussed as needed. sorbent, which is the primary determinant of the efficiency of sorption-based separations, the sizes of the gas molecules and the microstructure of the membrane play a crucial role in the separation performance. Membrane-based gas separations have been effectively used in many industrial-scale processes, such as H2/CO, H2/CH4, and H2/N2 separations, N2 separation from air, and natural gas sweetening, and may potentially be applicable for use in CO2 capture under the scenarios mentioned previously, namely, post-combustion CO2 capture (CO2/N2 separation), precombustion CO2 capture (CO2/H2 separation), and oxy-fuel combustion (O2/N2 separation). Although membrane-based separations of these pairs of gases have previously been demonstrated, the membranes must be further improved to function effectively under the specific operating conditions required for these CO2 capture processes. As mentioned in previous sections, the separations are often also complicated by the presence of minor components within the gas mixture. For example, in post-combustion CO2 capture, the partial pressure of CO2 within the feed stream is relatively low (∼0.15 bar) and the flue gas contains many other components, including H2O, SOx, and NOx. These relatively harsh conditions may limit the opportunity of membrane processes for application in post-combustion CO2 capture compared with sorption-based separation processes. However, recent work has suggested several membrane process designs for CO2 separation from flue gas mixtures along with detailed calculations on the energy requirements for each process.484 An excellent discussion on the desired membrane separation performance is also presented in the work. The performance of a membrane is usually described by its permeability (or permeance) and selectivity. The permeability, Pi, of gas component i in a given membrane can be calculated by the following expression in gas permeation experiments: Pi ¼ Ji  l Δpi 6.1. Continuous Films of Metal Organic Frameworks A growing number of continuous metal organic framework thin films have been fabricated for applications in a variety of areas, such as supported catalysts, molecular sensors, and gas separation membranes.486,487 In the context of gas separations, the studies to date have been limited to membranes with low surface areas, although significant advances have been made with regard to the investigation of their gas transport properties and the fabrication methods of these membranes on porous substrates. Initial efforts on the fabrication of metal organic framework continuous films have been made on nonporous supports such as uniformly functionalized organic self-assembled monolayers (SAM) and silicon wafers. For example, Fe-MIL-88B crystal layers were grown on a gold substrate functionalized with mercaptohexadecanoic acid by a solvothermal treatment with a synthesis mixture for MIL-53.488 In fact, two metal organic frameworks appeared after the reaction; the MIL-53 framework formed in the homogeneous solution, and MIL-88B framework was heterogeneously grown on the substrate. Uniformly distributed organic functional groups on the substrate allowed metal organic framework crystals to grow in an oriented fashion. In other studies, thin MIL-89489 and MIL-101490 layers were deposited on silicon wafers by a dip-coating method using a metal organic framework nanoparticle colloidal solution. Continuous metal organic framework films, however, should be fabricated on porous supports if they are to be used as gas separation membranes. The synthesis of layers of MOF-5 on a porous substrate has been accomplished by microwave heating,491 wherein a conductive layer such as graphite was deposited on a porous alumina substrate followed by selective and rapid formation of crystals of the metal organic framework on the coated surface. However, upon close examination of the morphology of the resulting films, it was determined that the metal organic framework layers were not continuous and were therefore not suitable for applications in gas separations. Nevertheless, continuous films were obtained through the conventional solvothermal method, and the resulting membranes deposited on an alumina support showed Knudsen diffusion behavior (H2/CO2 selectivity of 4.7) for all gases studied.492 Films of MOF-5 were fabricated using the seeded growth method assisted by microwave heating, and these also exhibited gas transport behavior consistent with a Knudsentype diffusion mechanism.493 Note that the specific gas transport mechanism observed for a given continuous metal organic framework film is derived predominantly from the structural features of the framework. Here, the large pore openings and channel dimensions (>10 Å in diameter) within MOF-5 presumably leads to a relatively low contribution of collisions of gas molecules with the pore walls, resulting in a Knudsen-type diffusion mechanism. However, one complicating factor in the analysis was the fact that intercrystalline boundary defects could not be excluded. The presence of such defects can lead to a significant contribution to the overall transport mechanism via ð11Þ where Ji is the molar flux of component i, l is the thickness of the membrane, and Δpi is the partial pressure difference across the membrane. The selectivity or permselectivity of a gas pair, αi/j, which is a measure of separation efficiency, can be expressed as the ratio of gas permeabilities: αi=j ¼ Pi Pj ð12Þ Note that when membranes have asymmetric structures, the permeance is used more frequently than the permeability: permeancei ¼ Ji Δpi ð13Þ Owing to their well-defined pore structures, the values of these parameters can be higher within crystalline microporous materials compared with polymeric membranes, where the gas transport is governed by a solution-diffusion behavior. Analogously, zeolite membranes with various topologies have demonstrated a significantly enhanced separation performance when compared with polymeric membranes. The high surface area and tunable pore functionalities available in metal organic frameworks presents an opportunity to further enhance the performance of membrane materials, although their use within membranes is still in a very preliminary stage.485 In this section, we discuss membranes containing metal organic frameworks either as a 763 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW diffusion of the gas molecules between the individual MOF-5 crystallites within the membrane, although such macroscopic defects were not observed using scanning electron microscopy (SEM). An HKUST-1 membrane has been fabricated on a copper surface via the hydrothermal method.494 The resulting membrane exhibited a higher H2 permeance than conventional zeolite membranes, with selectivity of ∼7 for H2 over CO2 at 25 °C and 1 bar of feed pressure. This can be attributed to the more rapid diffusion of H2 through the structure compared with CO2, which displays a relatively high affinity for the pore surface (particularly the Cu2+ sites) due to its greater quadrupole moment and polarizability. For the foregoing examples, the pore diameters within the metal organic framework structures are significantly larger than the kinetic diameters of the gas molecules, too large for size exclusion-based gas separation to be realized. However, a number of frameworks with smaller pores have been utilized in membranes to demonstrate size-based separations. For example, a continuous membrane has been fabricated from Cu(hfipbb)(H2hfipbb)0.5, which exhibits narrow one-dimensional channels with small pore windows (3.5  3.2 Å2 in diameter), comparable to those in small-pore zeolites, such as LTA, CHA, and DDR.495 Here, although continuous films could not be formed directly on bare alumina supports using solvothermal methods, submicrometersized crystals of the framework could be anchored to the surface after treatment of the substrate with a polyethyleneimine coating. The membrane was then fabricated by secondary growth from the seeded support and showed good gas selectivity, especially at high temperature (190 °C), with a selectivity of ca. 6 for CO2 over N2, and ca. 23 for H2 over N2 at 1 bar of feed pressure. However, the observed permeance within the membrane was very low, presumably as a result of the incorrect orientation of the one-dimensional channels of the structure with respect to the membrane. The gas transport properties of several continuous films of zeolitic imidazolate frameworks (ZIFs), including ZIF-8,496 498 ZIF-7,499 501 ZIF-22,502 and ZIF-90,503 were recently reported. Initially, films composed of ZIF-8 were fabricated on titania supports through a microwave-assisted solvothermal synthesis.496 As shown in Figure 33, a well-intergrown layer of ZIF-8 crystals was produced on the top of the support. The successful formation of continuous ZIF-8 layers was also visualized with a sharp transition between Zn and Ti signals in an energy-dispersive X-ray spectroscopy (EDX) map. The gas separation properties of the resultant membranes were investigated for various gases. Surprisingly, despite the pore diameter of 3.4 Å for ZIF-8 falling between the kinetic diameters of CO2 and CH4, the membrane showed a very low selectivity for CO2 over CH4 (α = 4 5) at 25 °C and 1.1 2 bar. The membrane also showed limited selectivity for CO2 over N2, and the selectivity for H2 over CO2 of ca. 4.5 is close to the Knudsen selectivity, although the membranes did show a higher H2 permeance compared with the Cu(hfipbb)(H2hfipbb)0.5 membranes. Continuous membranes of ZIF-7, which exhibits a pore diameter of approximately 3 Å, were fabricated by seeding of the alumina support with nanocrystals of the framework, followed by microwave-assisted solvothermal growth.499 Polyethyleneimine was added to the aqueous seed solution not only to stabilize the nanoparticles in the suspension but also to enhance the linkage between the seeds and the support through hydrogen bonding interactions. As shown in Figure 34, a well-intergrown Figure 33. SEM image (left) and corresponding EDX mapping (right) of a cross section of a ZIF-8 membrane. In the EDX map, orange and cyan represent Zn and Ti, respectively, clearly showing the clean formation of the ZIF-8 layer on the titania support. Reproduced with permission from ref 496. Copyright 2009 American Chemical Society. Figure 34. (a) Top down and (b) cross-section SEM images of a ZIF-7 membrane. (c) Cross-section SEM image of the ZIF-7 membrane (left) and the same image overlaid with an EDX mapping (right). In the EDX mapping, orange and cyan represent Zn and Ti, respectively, and the results clearly show the clean formation of the ZIF-7 layer on the alumina support. Reproduced with permission from ref 500. Copyright 2010 Wiley-VCH. ZIF-7 crystal layer was deposited on the support. Due to the thin nanoseed layer and the short secondary growth time during which microwave heating was performed, a ZIF-7 membrane with a thickness of just 1.5 μm, representing one of the thinnest metal organic framework-based membranes reported to date, could be readily prepared. This membrane showed relatively modest selectivities of 6.7 (single-component) and 6.5 (mixedcomponent) for H2 over CO2 at 200 °C and 1 bar. Here, it was hypothesized that the permeance of molecules with a diameter greater than the pore diameter, namely, CO2, N2, and CH4, is due to a contribution arising from non-size-selective mass transport through the imperfect sealing of the membrane or through the grain boundaries of the polycrystalline ZIF-7 film. Indeed, in a more detailed gas transport study within ZIF-7 membranes, an improved selectivity for H2 over CO2 of 13.6 was observed from an equimolar binary mixture at 220 °C and 1 bar.501 Significantly, the gas separation performance of the membrane did not deteriorate in the presence of steam, which is an important consideration for applications in post-combustion CO2 capture. In a more recent study, membranes of ZIF-22, which possesses the zeolite LTA topology with 3 Å pore windows, were fabricated 764 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW displaying selectivity for H2 over CO2 of 7.3 at 200 °C and 1 bar. However, permeation of molecules with a kinetic diameter larger than the crystallographic pore size of 3.5 Å led to low selectivities for CO2 over N2 and CO2 over CH4. Membranes comprising ZIF materials have also been reported using a variety of other substrate types. For example, a ZIF-8 film was deposited on a tubular porous alumina support seeded with ZIF-8 nanoparticles of ∼45 nm diameter.504 The resulting membrane showed very high CO2 permeance [(1.7 2.5)  10 5 mol/m2 3 s 3 Pa] with a selectivity for CO2 over CH4 of 4 7 at 22 °C and 1.4 bar. Another method for ZIF membrane synthesis involves modification of the surface of the support with the organic ligand of the metal organic framework of interest followed by conventional solvothermal synthesis.505 Here, it has been found that performing the synthesis at an elevated pH can play a crucial role in forming dense, well-intergrown membranes. In the presence of sodium formate, the imidazole units on the surface of the ZIF crystals are thought to be fully deprotonated due to the increase in pH (see Figure 35), resulting in isotropic growth in all crystal directions to yield large, high-quality crystals. In contrast, at low pH, random crystal branching and growth were observed, presumably as a result of only partial deprotonation of the surface imidazole groups. The permselectivities of the resulting ZIF-8 membranes were 11.6 for H2 over N2 and 3.9 for H2 over CO2 at 25 °C and 1 bar. More recently, ZIF-8 films were prepared on a flexible nylon substrate via a contra-diffusion of metal and imidazole linker solutions at room temperature.506 Here, two reagent solutions were initially separated by a porous nylon sheet within a U-tube, and then, following diffusion of the reagents through the nylon barrier, ZIF-8 crystals formed on both sides of the nylon sheet at the interfaces where the reagents combine. Unfortunately, the resulting membranes exhibited a selectivity for H2 over N2 of just 4, suggesting that their quality is relatively low compared with those fabricated using other methods of growth. Table 11 summarizes the CO2 separation performance of all of the continuous-film membranes formed from metal organic frameworks reported to date. Although some membranes exhibit very high gas permeabilities, significant improvements in the selectivities are required if they are to be adapted for use in a viable CO2 capture process. Significantly, the performance in the context of H2/CO2 selectivity is generally just slightly improved over the selectivity that would be expected from a Knudsen-type diffusion mechanism. Furthermore, metal organic framework membranes capable of performing CO2/N2 and O2/N2 separations have not yet been reported. However, the numerous examples of high-quality continuous-film membranes formed from zeolites and exhibiting attractive separation performances highlights the crucial need for the development of improved preparative methods for metal organic framework membranes. In particular, advances in control over the microstructure of the membranes, such as the minimization of intercrystalline grain boundary defects507,508 and control of the orientation of the crystallites,507,509 are important for the realization of highperformance membranes for CO2 capture applications. The formation of high-quality films is also crucial for fully characterizing the performance of metal organic frameworkbased membranes. Efforts to better characterize the intrinsic capabilities of metal organic frameworks in membrane-based separation processes have also begun to emerge. Recently, the gas transport Scheme 2. Surface Modification of Al2O3 with 3-Aminopropyltriethoxysilane Moieties, Which Can Then Be Used in Covalently Grafting Imidazolate Functionalities for Anchoring ZIF-90 Crystals onto the Surface503 on porous titania supports treated with 3-aminopropyltriethoxysilane (APTES).502 Here, it was speculated that the 3-aminopropylsilyl groups could coordinate to the exposed Zn2+ centers on the surface of the nanocrystals, enabling them to be directly tethered to the substrate surface, where they could then act as seeds for further film growth. However, in a similar case to the ZIF-7 films, the permeation of molecules larger than the pore diameter of 3 Å was large, and the membrane showed only a modest selectivity for H2 over CO2 of 7.2 at 50 °C and 0.5 bar. When the feed pressure was increased to 1 bar, the H2/CO2 selectivity decreased to 5.1. Continuous films of ZIF-90, which is isostructural to ZIF-8 but with aldehyde groups on the organic bridging unit, have also been fabricated on an alumina support treated with APTES.503 In this work, the aldehyde groups within the organic linker were utilized for the covalent anchoring of the framework onto the surface through a reaction with the aminopropyl group installed on the support (see Scheme 2). Subsequent nucleation and film growth of ZIF-90 about these sites resulted in a high-quality membrane 765 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Figure 35. A schematic of the proposed role of sodium formate in promoting the growth of large ZIF-8 particles through complete deprotonation of the surface imidazole functionalities. Top-view SEM images of ZIF-8 films after secondary growth are also shown (a) in the presence of sodium formate and (b) without addition of sodium formate. The larger crystal sizes in panel a result in a more dense membrane, which is essential for good performance in gas separation applications. Reproduced with permission from ref 505. Copyright 2010 American Chemical Society. Table 11. CO2 Separation Performance of Metal Organic Framework Continuous Films metal organic framework gas permeance (mol/m2 3 s 3 Pa) 8 Cu(hfipbb)-(H2hfipbb)0.5 H2, 2  10 Cu3(BTC)2 H2, (1.0 1.3)  10 Zn4O(BDC)3 H2, 4.7  10 SIM-1 H2, 8  10 ZIF-7 H2, (7.5 8.0)  10 6 6 8 8 selectivity testing conditions ref H2/CO2, ∼4 25 190 °C, 1 bar 495 H2/CO2, ∼7 25 °C, 1 bar 494 H2/CO2, 4.7 25 °C, 1 bar 492 H2/CO2, ∼2.5 50 °C, 0.5 3.0 bar 512 H2/CO2, ∼7 200 °C, 1 bar 500 H2/CO2, 13 220 °C, 1 bar 501 H2/CO2, 4.5 CO2/CH4, 4 7 25 °C, 1.1 2.0 bar 22 °C, 1.4 bar 496 504 7 H2/CO2, 7.2 50 °C, 0.5 bar 502 7 H2/CO2, 7.3 200 °C, 1 bar 503 8 ZIF-7 H2, 4.5  10 ZIF-8 ZIF-8 H2, 5.1  10 8 CO2, 1.7 2.5  10 ZIF-22 H2, 2.0  10 ZIF-90 H2, 2.5  10 5 properties have been measured through a single crystal of Cu2(bza)4(pyz) (bza = benzoate; pyz = pyrazine), the structure of which exhibits a pore size of about 4 Å.510 The high dependence of the gas transport properties on the orientation of the metal organic framework crystal (see Figure 36) illustrates the importance of controlling the growth direction within polycrystalline films. In addition, the gas selectivities observed for this particular single-crystal membrane are greater than for any other metal organic framework membrane reported to date, with a CO2/N2 selectivity of 14, and a CO2/CH4 selectivity of 25. Interestingly, CO2 is more permeable than H2 in this structure, suggesting an adsorption-selective mechanism rather than a simple diffusion-selective mechanism. Such studies clearly indicate that the gas transport performance of metal organic framework membranes can indeed be high, but only upon incorporation within high-quality structures. For the preparation of high-performance membranes, the metal organic framework should be wisely chosen to suit the target separations. To this end, computational studies on mass Figure 36. Single-component gas permeabilities of a Cu2(bza)4(pyz) single-crystal membrane when the crystal orientation is controlled such that the one-dimensional channels within the structure of the framework are aligned (red) and perpendicular (turquoise) with respect to the direction of the incoming gas flow. Reproduced with permission from ref 510. Copyright 2010 American Chemical Society. transport properties in metal organic frameworks can be used to predict the performance of the resultant membranes. For 766 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Figure 37. Schematic illustration of mixed-matrix membranes (a) incorporating metal organic framework crystallites within a dense film and (b) in an asymmetric membrane geometry. applications in size-exclusion based gas separations. The gas permeability measured at 35 °C and 1 bar was found to decrease slightly as a function of the amount of metal organic framework loaded within the polymer matrix, although membranes containing 5 wt % of the framework did show an enhanced selectivity of ca. 200 for H2 over CH4 versus ca. 50 for the pure polymeric membrane. Unfortunately, the CO2 transport properties of this system were not reported. Mixed-matrix membranes have also been studied with varying amounts of HKUST-1 incorporated within two different polymers: a rubbery polydimethylsiloxane (PDMS) and a glassy polysulfone.514 The polysulfone membrane showed a gradual increase in CO2 permeability with increasing HKUST-1 loading, up to 10 wt %. Membranes containing 5 wt % of HKUST-1 exhibited further enhanced selectivity of ca. 25 for CO2 over N2, as compared with ca. 20 for pure polysulfone membranes. The selectivity dropped significantly when the framework loading was increased to 10 wt %. Meanwhile, a 10 wt % HKUST-1/ PDMS membrane showed a substantial enhancement in CO2 permeability from 2500 to 3000 barrers, with a modest increase in the selectivity from 8.2 to 8.8. Here, again, the performance was not improved when the metal organic framework loading was further increased.514 Mixed-matrix membranes containing CuSiF6(bpy)2 and MOF-5 within a commercial polyimide (Matrimid) have been fabricated and tested.515 Gas permeation measurements performed at 35 °C and 2 bar revealed that incorporating the copper-based framework, which exhibits a structure with 8 Å pore windows, within the polymer increased the CO2 permeability but decreased the selectivity for CO2 over N2. In contrast, the selectivity for CH4 over N2 was enhanced. This can be rationalized on the basis that the framework has a high affinity for CH4, presumably due to its large surface area and relatively large pore windows. The Matrimid membranes incorporating MOF-5 showed increased CO2 permeability with a modest increase or no change in gas selectivity, according to pure gas permeation tests conducted at 35 °C and 2 bar.516 As the best example, in a membrane containing 30 wt % MOF-5, the CO2 permeability was almost doubled with a 10% increase in the selectivity for CO2 over N2 compared with the pure polymer membrane. Disappointingly, the mixed gas permeation tests for this membrane formulation revealed H2/CO2 and CO2/CH4 selectivities lower than the corresponding values of the pure polymeric membrane. One of the major challenges for mixed-matrix membrane studies is the issue of nonideality at the interface between the dispersed microporous crystals phase and the polymer matrix, particularly when glassy polymers are employed. To overcome this problem, a mixed-matrix membrane based on a low-Tg (glass transition temperature) polymer, polyvinylacetate (PVAc) was fabricated in an attempt to achieve a defect-free morphology.517 Mixed-matrix membranes containing example, a recent study predicted an excellent CO2 separation performance in a Mg2(dobdc) membrane.511 The permselectivity of 200 estimated for a CO2/H2 mixture at 27 °C and 10 bar was higher than that of any other metal organic framework studied, as well as the experimental result for the SAPO-34 membrane, which possesses the zeolite CHA topology with 3.8 Å pore window. The selectivity estimated for the CO2/N2 mixture was also higher than 40 in this material. This outstanding performance arises from the chemical and structural properties of Mg2(dobdc), in which, as discussed in section 3.2.3, a large concentration of open metal sites allows CO2 to be selectively adsorbed onto the pore channels over other species. Then, in the one-dimensional channels of the framework, CO2 molecules bound to pore walls can effectively slow the more mobile partner species (H2 and N2), resulting in a high CO2 permselectivity in the mixture. The CO2 permeance in the Mg2(dobdc) membrane was also predicted to be much higher than that of the SAPO-34 membrane, owing to its larger pore dimension (11 Å), further emphasizing the outstanding performance of Mg2(dobdc) and its potential as a membrane material for applications in CO2 capture. 6.2. Mixed-Matrix Membranes An alternative approach to the formation of metal organic framework-based separation membranes is through the incorporation of crystals into a polymeric membrane matrix, as depicted in Figure 37. One key advantage of such “mixed-matrix” membranes is the ability to combine the processability and mechanical stability of polymers with the excellent gas separation properties of crystalline, microporous solids. Zeolite molecular sieves have been most widely employed for this purpose so far, but reports of studies with mixed-matrix membranes containing metal organic frameworks have also begun to appear.15 The incorporation of metal organic frameworks is particularly attractive due to their high surface areas, adjustable pore dimensions, and tunable surface functionality, which might enable the fabrication of membranes exhibiting properties that are optimized for a specific separation. Furthermore, the high processability originating from the polymer phase could allow large-scale fabrication of high-quality membranes for industrial applications using conventional methods. Similar to the continuous metal organic framework membranes described in section 6.1, their mixed-matrix counterparts are still in the early stages of investigation, and many improvements are required in the selection of materials and development of fabrication methods in order to realize high-performance membranes. Composite membranes incorporating the three-dimensional metal organic framework Cu2(PF6)(NO3)(bpy)4 3 2PF6 3 2H2O confined within an amorphous glassy polysulfone polymer have been investigated.513 The very narrow channels within this framework structure, measuring 4  3 Å2 in the crystallographic b-direction and 3  3 Å2 in the c-direction, should be suitable for 767 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Cu(BDC) showed enhanced separation performance compared with the pure PVAc membrane when tested at 35 °C and 4.5 bar. Membranes containing 15 wt % of Cu(BDC) exhibited a 34% increase in the CO2 permeability and a 10% increase in the selectivity for CO2 over N2. This rather modest gas separation enhancement can be attributed to both a molecular sieving effect from the relatively small pores of the framework (5.2 Å pore windows) and improved adhesion between the metal organic framework and polymer phases. A number of ZIF-based mixed-matrix membranes have also been reported. In particular, the CO2 transport properties of ZIF8/polysulfone mixed-matrix membranes have been investigated via gas permeation measurements and pulsed-field gradient NMR techniques.518 The CO2 self-diffusion coefficient measured by NMR spectroscopy at 25 °C increased with ZIF-8 content, from 2.1  10 8 cm2/s for pristine membranes to 9.3  10 8 cm2/s for membranes containing 30 wt % ZIF-8. Gas permeation tests confirmed the significantly improved CO2 permeability for the mixed-matrix membrane. Detailed analysis further revealed that the ZIF-8 particles also contribute greatly to the CO2 permeation by increasing the gas solubility within the composite membranes. Additionally, ZIF-8/Matrimid mixedmatrix membranes were tested with various single-component and mixed gases at 35 °C and 2.7 bar.519 Gas permeability increased with the loading of ZIF-8 up to a level of 40 wt %. Noticeable increases in selectivities, as evaluated from both single-component and mixed gas permeation, were observed at a loading level of 50 wt % for gas pairs including small molecules, such as H2/CH4, CO2/CH4, H2/C3H8, and CO2/C3H8, suggesting that the ZIF-8 phase may act as a molecular sieve for smaller gas molecules. Membranes comprising ZIF-90 dispersed within various polyimides have also been investigated.520 As shown in Figure 38, the ZIF-90 crystals were evenly distributed within the polymer matrix and interfacial voids were not observed by SEM. Gas permeation tests revealed that when less permeable polymers such as Ultem and Matrimid were used as the matrix, the membrane displayed an increase in the CO2 permeability with no change in the selectivity for CO2 over CH4. However, an enhancement in both permeability and selectivity was observed when a highly permeable polyimide 6FDA DAM (6FDA = 2,2bis(3,4-carboxyphenyl)hexafluoropropane dianhydride, DAM = diaminomesitylene) was employed as the matrix. This result highlights the importance of a careful selection of both the polymer and metal organic framework phases based on gas permeabilities, such that the performance of the resulting mixedmatrix membrane is optimized for a given separation. Indeed, one of the most promising membranes displayed excellent performance for post-combustion CO2 capture; exhibiting a CO2 permeability of ca. 700 barrers with a selectivity for CO2 over N2 of 22 at 25 °C and 2 bar. Most recently, in order to improve the CO2 separation performance of polymeric membrane, an amino-functionalized metal organic framework, NH2-MIL-53, was incorporated into polysulfone.439 Presumably due to the hydrogen bonding interactions between the amine functionalities within the metal organic framework crystal and the sulfone groups of the polymer, high NH2-MIL-53 loadings of up to 40 wt % were achieved. However, the best CO2/CH4 separation performance was observed in 25 wt % NH2-MIL-53 loaded membranes. In the gas permeation study at 35 °C and 3 bar, although there was no significant improvement in CO2 permeability, Figure 38. SEM images of cross-sections of mixed-matrix membranes incorporating ZIF-90 within (a) Ultem, (b) Matrimid, and (c) 6FDA DAM. Reproduced with permission from ref 520. Copyright 2010 Wiley-VCH. CO 2 /CH4 selectivity was almost doubled in the mixedmatrix membranes as compared with pure polysulfone membranes. Even higher selectivity (ca. 110) was observed at 10 °C and 10 bar, although the CO 2 permeability decreased significantly. The aforementioned mixed-matrix membranes were all measured in a dense film form to investigate their intrinsic gas transport properties. However, such membranes should ideally be fabricated in an asymmetric form in which a thin active surface layer is supported by a porous sublayer to minimize the mass transport resistance (see Figure 37b). Only a very limited number of studies investigating the gas separation performance of asymmetric mixed-matrix membranes containing metal organic frameworks have been reported to date. Asymmetric HKUST-1/Matrimid mixed-matrix membranes have been fabricated via the phase inversion method.521 Here, to seal the defects within the membrane, the metal organic framework/polymer thin surface layer was coated with the highly permeable polymer PDMS. The resultant membrane showed a substantial increase in CO2 permeance with modest enhancement in the selectivity for CO2 over N2 compared with pure Matrimid membranes. Asymmetric membranes with HKUST-1 crystallites embedded within a polyimide matrix have been prepared in a hollow fiber form, a capillary geometry with a small diameter.522 These displayed an enhancement in both the H2 permeability and selectivity for H2 over CO2, although the CO2/N2 separation performance was not improved by incorporating the metal organic framework into the polymer. 768 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW Table 12. CO2 Separation Performance of Metal Organic Framework Mixed-Matrix Membranes gas permeability (barrers)a MOF and polymer selectivity testing conditions ref Cu(BDC), PVAc CO2, 3.3 CO2/N2, 35 35 °C, 4.5 bar 517 HKUST-1, PDMS CO2, 3000 CO2/N2, 9 not reported 514 HKUST-1, polysulfone CO2, 6 8 CO2/N2, 25 not reported 514 CuSiF6(bpy)2, Matrimid H2, 17 26 H2/CO2, 2 35 °C, 2 bar 515 CO2 8 15 CO2/N2, 30 33 H2, 30 54 H2/CO2, 2 3, 35 °C, 2 bar 516 CO2, 11 20 CO2/N2, 35 40 ZIF-8, Matrimid H2, 29 71 CO2, 9 24 H2/CO2, 3 4 CO2/N2, 23 30 35 °C, 2.7 bar 519 ZIF-90, 6FDA-DAM CO2, 600 800 CO2/N2, 22 25 °C, 2 bar 520 MOF-5, Matrimid CO2/CH4, 37 HKUST-1, Matrimidb b HKUST-1, polyimide a 1 barrer = 3.348  10 16 CO2, 12 18 GPUc CO2/N2, 23 27 35 °C, 10 bar 521 H2, 876 GPU H2/CO2, 10 25 °C, 2 bar 522 CO2, 88 GPU CO2/N2, 9 mol 3 m/(m2 3 s 3 pa). b Asymmetric membranes. c Gas permeation unit, 10 Figure 39. CO2/N2 separation performance of metal organic framework mixed-matrix membranes. The solid line represents the upper bound of polymeric membranes, as determined in 2008.524 ZIF-8/ Matrimid from ref 519; ZIF-90/6FDA-DAM from ref 520; MOF-5/ Matrimid from ref 516; Cu(BDC)/PVAc from ref 517; CuSiF6(bpy)2/ Matrimid from ref 515; HKUST-1/PDMS from ref 514; HKUST-1/ polysulfone from ref 514. 6 cm3(STP)/(cm2 3 s 3 cm Hg). Figure 40. Predicted CO2/CH4 separation performance of hypothetical mixed-matrix membranes. Squares represent hypothetical pure polymer membranes on the upper bound limit (Robeson line). Triangles and circles show the predictions for a mixed-matrix membrane with a 30% volume fraction of IRMOF-1 (MOF-5) and Cu(hfipbb)(H2hfipbb)0.5, respectively. The performance of several well-known polymer membranes are also shown. Reproduced with permission from ref 523. Copyright 2010 The Royal Society of Chemistry. The CO2 separation performance of metal organic framework-based mixed-matrix membranes is summarized in Table 12. Due to the properties of the polymer matrix, these membranes generally exhibit enhanced gas selectivity compared with the pure metal organic framework films, especially in the case of CO2/N2 separations. Figure 39 shows the CO2/ N2 separation performance of the mixed-matrix membranes reported to date. Although many membranes exhibit enhanced characteristics compared with pristine polymers, no metal organic framework mixed-matrix membrane in which the performance exceeds that of the best polymers has been demonstrated to date. As mentioned above, the preparation of high-performance membranes relies on the careful selection of both the polymer and the metal organic framework phases.520,523 This is also highlighted in a computational study on metal organic framework mixed-matrix membranes.523 As shown in Figure 40, CO2/CH4 selectivity cannot be improved compared with pristine polymers when MOF-5, possessing much larger pores than the kinetic diameters of both CO2 and CH4, is incorporated into the polymer matrix. In contrast, mixed-matrix membranes comprised of a small-pore metal organic frameworks, such as Cu(hfipbb)(H2hfipbb)0.5, can show higher gas separation performance than pure polymeric membranes. The properties of the resulting mixed-matrix membranes can also be tuned depending on the selection of the polymer. In particular, when the gas permeabilities of both phases are well-matched, the enhancement in gas separation performance can be maximized. In addition, a high degree of control over the interfacial morphology between the metal organic 769 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW metal organic frameworks now has emerged. Nevertheless, there remains a need for a broader range of frameworks to be examined in order to identify the materials that provide the best properties for application in a real-world capture process. Similar to post-combustion capture, there is a need for the effect of the minor components of the flue gas (predominantly CO and CH4) to be fully elucidated, although these trace impurities are not expected to significantly interfere with the performance of the materials owing to their relatively weak interactions with the surfaces of metal organic frameworks. Moreover, the practical aspects of employing a pressure swing adsorption process, such as the desorption of CO2 by purge, have not yet been investigated in detail. More thorough investigations of the complete process under the conditions relevant for pre-combustion CO2 will allow the prospects of employing these materials in an industrial setting to be more accurately evaluated. Oxy-fuel combustion has emerged as an application in which the chemical tunability of metal organic frameworks may allow high-performance materials to be prepared. The ability of the exposed electron-donating metal centers within Fe2(dobdc) to bind O2 preferentially via charge transfer interactions has led to a high O2/N2 selectivity, but the low temperatures at which the binding can be reversed is a significant drawback of the material. At room temperature, the O2 adsorption is irreversible owing to a two-electron transfer process and the concomitant formation of a peroxide (O22 ) species bound to the metal centers. An enhanced control over the electronic properties of the exposed metal cation sites would allow the electron transfer to be limited to a one-electron process, which has been shown to permit the reversible binding of O2 at high capacity. The generation of new metal organic frameworks bearing exposed redox-active metals (such as Cr2+, Mn2+, Fe2+, Co2+, or Cu+) and the fine-tuning of the electron density available at these adsorption sites through substitution of the organic backbone of the materials are expected to afford new materials that reversibly bind O2 near ambient temperatures with high selectivity and working capacity. The large-scale implementation of metal organic frameworks for CO2 capture will also inevitably require the consideration of a number of other aspects in addition to the adsorptive properties of the materials. The performance of the materials should be evaluated in the context of engineering process models that take performance parameters, such as the working capacity and selectivity, as input to allow optimized working conditions to be developed for each adsorbent. Lifecycle analyses should be performed in order to ascertain the feasibility of bulk preparations of the most promising metal organic frameworks, taking into account the many practical considerations for large-scale synthesis, including the supply chain of the precursor materials, cost, environmental impact of waste products, and recycling following use of the adsorbent. Before such detailed analyses are performed, however, the synthetic challenge of preparing robust new materials that are truly well-suited to each capture scenario must be addressed. Owing to the large number of possible metal organic framework structures that can be assembled, the use of computational and theoretical approaches may assist the synthetic efforts by identifying new target materials predicted to display the appropriate properties. The tremendous progress already made in improving the properties of metal organic frameworks is indeed promising for real-world applications, and we remain optimistic that materials capable of serving as next-generation CO2 capture framework and the polymer is also a prerequisite for the fabrication of high-performance mixed-matrix membranes. Many efforts have been made to obtain improved morphologies, including particle/polymer adhesion and the uniform distribution of particles. The methodologies developed for zeolites or other molecular sieves do not necessarily apply to metal organic frameworks, but it is likely that new and improved fabrication techniques customized specifically for these materials will soon afford membranes with enhanced gas separation properties. 7. CONCLUDING REMARKS AND OUTLOOK The foregoing sections have described the recent progress made in the investigation of metal organic frameworks as potential new solid adsorbents to be used within CO2 capture systems. Indeed, tremendous CO2 adsorption capacities have been demonstrated in the highest surface area materials, and high adsorptive selectivities have also begun to emerge in materials furnished with functionalized surfaces. With regard to the prospects of creating new materials suitable for real-world applications, the high degree of control over the structural and chemical features of metal organic frameworks is particularly promising for optimization of their properties, not only for the type of CO2 capture to be performed but also for the specific composition of the flue gas of a particular power plant. Such a precise tuning of the material characteristics is expected to result in considerable improvements in the sorbent performance, leading to reduced energy requirements for the capture process compared with current technologies. In this section, we briefly outline the main issues that need to be addressed in order to achieve next-generation materials capable of fulfilling the criteria required for the gas separations relevant to post-combustion CO2 capture, pre-combustion CO2 capture, and oxy-fuel combustion. In the area of post-combustion CO2 capture, the primary need is for more chemically and thermally robust materials capable of withstanding the high levels of water present in the flue gas stream, while also tuning the temperature required for regeneration. Increasing the strength of the metal ligand bonds through the incorporation of high-valent metal cations (e.g., Al3+ and Ti4+) or more strongly binding organic ligands (e.g., pyrazolates and triazolates) is one approach by which this may be achieved. Furthermore, although most of the adsorption studies to date have evaluated the performance of materials through single-component CO2 and N2 adsorption isotherms or breakthrough experiments using a CO2/N2 mixed gas, the impact of the presence of water and other minor components in the flue gas is an aspect where a greater understanding is urgently needed. In addition to affecting the separation performance, the impurity components (O2, CO, SOx, NOx) may have significant consequences in terms of the long-term stability of the materials, as well as the energy requirements for regeneration. Thus, the study of metal organic frameworks under conditions that simulate realistic working conditions (gas composition, temperature, and pressure) is essential for fully evaluating the performance of a given material. For pre-combustion CO2 capture, metal organic frameworks have recently been demonstrated to show enhanced CO2/H2 selectivities and working capacities over conventional materials, such as activated carbons and zeolites. Although only a small number of compounds have been experimentally studied to date, a significant body of theoretical work examining this separation in 770 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW adsorbents will be discovered through continued investigations in the field. AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]. Author Contributions † These authors contributed equally to this work. BIOGRAPHIES Jarad A. Mason was born in 1987 in North Carolina. He received his B.A. (2009) and M.S. (2009) in Chemistry at the University of Pennsylvania, where he studied polymeric precursors to ultra-hightemperature ceramics under the supervision of Prof. Larry G. Sneddon. He is currently a graduate student in Jeffrey Long’s research group at the University of California, Berkeley. His research focuses on the synthesis of highly robust metal organic frameworks for gas storage and separation applications. Kenji Sumida was born in 1982 in New Zealand and obtained his B.Sc. in Computer Science in 2002 and his M.Sc. degree in Chemistry in 2007 at the University of Auckland under the supervision of Prof. Penny J. Brothers. He joined Jeffrey Long’s research group at the University of California, Berkeley, in 2007, where his work focuses on the development of highthroughput methodologies for the discovery of novel metal organic frameworks for gas storage and molecular separation applications. Thomas M. McDonald was born and raised in Denver, Colorado. In 2007, he received a B.A. in Chemistry from Northwestern University, where he worked under the guidance of Prof. Teri Odom and Prof. Ken Poeppelmeier. He joined the research laboratory of Jeffrey Long in 2008 and is currently researching novel porous materials for gas separations and storage. David L. Rogow was born in 1976 in Las Vegas, Nevada. He received his B.S. (2005) in Chemistry at Portland State University, Portland, Oregon, and earned his Ph.D. (2010) at the University of California, Santa Cruz, in Chemistry and Biochemistry under the supervision of Prof. Scott R. J. Oliver. He is currently a postdoctoral fellow (NSF) in Jeffrey Long’s research group at the University of California, Berkeley. His research focuses on the synthesis of functionalized metal organic frameworks for carbon dioxide capture. Eric D. Bloch was born in 1983 in Wisconsin. He received his B.S. (2008) in Chemistry at the University of Wisconsin—Milwaukee and is currently a graduate student in Jeffrey Long’s research group at the 771 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW University of California, Berkeley. His research focuses on the synthesis of metal organic frameworks containing redox-active metal centers for gas separation applications. Ph.D. in Chemistry at Harvard University in 1995. Following postdoctoral work at Harvard and the University of California, Berkeley, he joined the faculty at Berkeley in 1997, where he is currently Professor of Chemistry. He is also a Faculty Senior Scientist in the Materials Sciences Division of Lawrence Berkeley National Laboratory. His research involves the synthesis of new inorganic clusters and solids, with emphasis on magnetic and microporous materials. ACKNOWLEDGMENT This research was funded through the Center for Gas Separations Relevant to Clean Energy Technologies, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Award No. DE-SC0001015. We thank Fulbright New Zealand for partial support of K.S., the NSF ACC-F Award No. 1042021 for support of D.L.R., and NSF for fellowship support of J.A.M. Zoey R. Herm was born in 1984 in Chicago, Illinois. She received her B.A. (2007) in Chemistry at Macalester College in St. Paul, Minnesota, where she studied group VI organometallic carbonyl complexes under the supervision of Prof. Paul J. Fischer. She is currently a Ph.D. student in Jeffrey Long’s research group at the University of California, Berkeley. Her research focuses on precombustion carbon dioxide capture in metal organic frameworks. LIST OF ABBREVIATIONS abtc 3,30 ,5,50 -azobenzenetetracarboxylate ACMP acetylene-mediated conjugated microporous polymers ADA adamantanediacetate ADC 4,40 -azobenzenedicarboxylate ade adenine almeIm 4-methylimidazole-5-carbaldehyde AZPY 4,40 -azo(bis)pyridine atz 3-amino-1,2,4-triazole BBC 4,40 ,400 -(benzene-1,3,5-triyl-tris(benzene-4,1-diyl))tribenzoate bbIm 5-bromobenzimidazole BBPDC 40 -tert-butyl-biphenyl-3,5-dicarboxylate bcphfp 2,20 -bis(4-carboxyphenyl)hexafluoropropane BDC 1,4-benzenedicarboxylate bdcppi N,N0 -bis(3,5-dicarboxyphenyl)pyromellitic diimide bdi 5,50 -(buta-1,3-diyne-1,4-diyl)diisophthalate BDoborDC 1,4-bis(1,12-dicarbonyl-closo-dodecaborane)benzene benzDHP 1,4-benzenedihydrogenphosphonate benzTB N,N,N0 ,N0 -benzidinetetrabenzoate bIm benzimidazole BME-bdc 2,5-bis(2-methoxyethoxy)benzene-1,4-dicarboxylate BPDC biphenyl-4,40 -dicarboxylate bpe 1,2-bis(4-pyridyl)ethane bpee 1,2-bis(4-pyridyl)ethylene bpetha 1,2-bis(4-pyridyl)ethane BPnDC benzophenone-4,40 -dicarboxylic acid bpp 1,3-bis(4-pyridyl)propane bpta 3,6-di(4-pyridyl)-1,2,4,5-tetrazine bptb 2,20 -biphenol-3,30 ,5,50 -tetrakis(4-benzoate) BPTC 3,30 ,5,50 -biphenyltetracarboxylic acid bpy 4,40 -bipyridine bpydc 2,20 -bipyridine-5,50 -dicarboxylate BTB 4,40 ,400 -benzene-1,3,5-triyl-tribenzoate BTC 1,3,5-benzenetriscarboxylate btdc 2,20 -bithiophene-5,50 -dicarboxylate BTE 4,40 ,400 -(benzene-1,3,5-triyl-tris(ethyne-2,1-diyl))tribenzoate btei 5,50 ,500 -benzene-1,3,5-triyltris(1-ethynyl-2-isophthalate) BTetB 4,40 ,400 ,4000 -benzene-1,2,4,5-tetrayltetrabenzoic acid Tae-Hyun Bae was born in 1977 in South Korea and received his B.S. (1999), M.S. (2001), and Ph.D. (2006) degrees in natural fiber science at Seoul National University. He earned his second Ph.D. (2010) in chemical engineering at Georgia Institute of Technology under the supervision of Prof. Christopher W. Jones and Prof. Sankar Nair. Currently, he is a postdoctoral fellow in Jeffrey Long’s research group, where he is working on CO2 separation with metal organic frameworks. Jeffrey R. Long was born in Rolla, Missouri, USA, in 1969. He received his B.S. in Chemistry from Cornell University in 1991 and 772 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW BTP bttb cbIm CDC 1,3,5-tris(1H-pyrazol-4-yl)benzene 4,40 ,400 ,4000 -benzene-1,2,4,5-tetrayl-tetrabenzoate 5-chlorobenzimidazole 1,12-dihydroxydicarbonyl-1,12-dicarba-closododecaborane CEbnbpy (R)-6,60 -dichloro-2,20 -diethoxy-1,10 -binaphthyl4,40 bipyridine cis-chdc 1,4-cyclohexanedicarboxylate CNBPDC 2,20 -dicyano-4,40 -biphenyldicarboxylic acid CNC 4-carboxycinnamate cnIm 4-cyanoimidazole CNT carbon nanotubes CPI 5-(4-carboxy-phenoxy)isophthalate CPOM [(4-carboxyphenyl)oxamethyl]methanoate cyamIm 4-aminoimidazole-5-carbonitrile DABCO triethylenediamine (1,4-diazabicyclo[2.2.2]octane) DBM dibenzoylmethanato DBNBVP (S)-2,20 -diethoxy-1,10 -binaphthyl-6,60 -bis(4-vinylpyridine) DBS 4,40 -dibenzoate-2,20 -sulfone dcIm 4,5-dichloroimidazole dhbc 2,5-dihydroxybenzoate dhbpc 4,40 -dihydroxybiphenyl-3-carboxylate DHT dihydroxyterephthalic acid diazBDC 5,50 -(1E)-1,2-diazinediylbis(1,3-benzendicarboxylate) DCTP 3,300 -dicarboxy-1,10 :40 ,100 -terphenyl dImb 1,4-di(1H-imidazol-4-yl)benzene diPyNI N,N0 -di-(4-pyridyl)-1,4,5,8-naphthal-enetetracarboxydiimide DMA N,N0 -dimethylacetamide dobdc 2,5-dioxido-1,4-benzenedicarboxylate doborDC 1,12-dihydroxy-carbonyl-1,12-dicarba-closododecaborane dpa 1,10 -biphenyl-2,20 -dicarboxylate dpe 1,2-di(4-pyridyl)ethylene DPG meso-1,2-bis(4-pyridyl)-1,2-ethanediol DPNI N,N0 -di-(4-pyridyl)-1,4,5,8-naphthalenetetracarboxydiimide dpntcd N,N-di(4-pyridyl)-1,4,5,8-naphthalenetetracarboxydiimide DPT 3,6-di-4-pyridyl-1,2,4,5-tetrazine dpt24 3-(2-pyridyl)-5-(4-pyridyl)-1,2,4-triazolate DSbpyDO 4,40 -disulfo-2,20 -bipyridine-N,N0 -dioxide dtp 2,3-di-1H-tetrazol-5-yl-pyrazine EBTC 1,10 -ethynebenzene-3,30 ,5,50 -tetracarboxylate eIm 2-ethylimidazole ELM elastic layer-structured MOF etz 2,5-diethyl-1,2,4-triazole FMA fumarate F-pymo 5-fluoropyrimidin-2-olate Gly-Ala glycylalanine hymeIm 4-hydroxymethyl-5-methylimidazole H4abtc 1,10 -azobenzene-3,30 ,5,50 -tetracarboxylic acid H4bdi 5,50 -(buta-1,3-diyne-1,4-diyl)diisophthalic acid biphenyl-3,40 ,5-tricarboxylate H3BPT H2bpydc 2,20 -bipyridine-5,50 -dicarboxylic acid H2BPDC 4,40 -biphenyldicarboxylic acid H4DHBP 1,4-dihydroxy-2,5-benzenediphosphonic acid 4,40 -(hexafluoroisopropylidene)-bis(benzoic acid) H2fipbb H3idc 4,5-imidazoledicarboxylic acid H4MTB methanetetrabenzoic acid H2obb 4,40 -oxybis(benzoic acid) Hoxonic H2ppt H3pzdc IBPDC IDC Im IN MAF MAMS mbIm mdpt24 Me4bpz mIm m-TATB nbIm NDC ndc nIm NTC ntei 4,6-dihydroxy-1,2,3-triazine-2-carboxylic acid 3-(2-phenol)-5-(4-pyridyl)-1,2,4-triazole 3,5-pyrazoledicarboxylic acid dimethyl-2,20 -diiodo-4,40 -biphenyldicarboxylate 2-methylimidazolate-4-amide-5-imidate imidazole isonicotinate metal azolate framework mesh-adjustable molecular sieves 5-methylbenzimidazole 3-(3-methyl-2-pyridyl)-5-(4-pyridyl)-1,2,4-triazolate 3,30 ,5,50 -tetramethyl-4,40 -bis(pyrazolate) 2-methylimidazole 3,30 ,300 -s-triazine-2,4,6-triyltribenzoate 5-nitrobenzimidazole 2,6-naphthalenedicarboxylate naphthalene-2,6-dicarboxylic acid 2-nitroimidazole naphthalene-1,4,5,8-tetracarboxylate 5,50 ,500 -(4,40 ,400 -nitrilotris(benzene-4,1-diyl)tris(ethyne-2,1-diyl))triisophthalate pba 4-(pyridin-4-yl)benzoate pbmp N,N-piperazinebismethylenephosphonate p-CDC 1,12-dihydroxycarbonyl-1,12-dicarba-closododecaborane PCN porous coordination network PDA phenylenediacetate pdc 3,5-pyridinedicarboxylate PDC pyrenedicarboxylic acid pegBTB 2,20 -penta(ethylene glycol)biphenyl-3,30 ,5,50 -tetrakis(4-benzoate) phen 1,10-phenanthroline pmc pyrimidine-5-carboxylate pmdc pyrimidine-4,6-dicarboxylate ptei 5,50 -((50 -(4-((3,5-dicarboxyphenyl)ethynyl)phenyl)[1,10 :30 ,100 -terphenyl]-4,400 -diyl)-bis(ethyne-2, 1-diyl))diisophthalate Pur purine py pyridine pydc 3,5-pyridinedicarboxylate pymo pyrimidinolate pyrdc pyridine-2,3-dicarboxylate pyz pyrazine pzdc 2,3-pyrazinedicarboxylate SCA sulfonatocalix[4]arene TATB 4,40 ,400 -s-triazine-2,4,6-triyl-tribenzoate TBA tetrabutylammonium TCEPEB 1,3,5-tris[(1,3-carboxylic acid-5-(4-ethynyl)phenyl))ethynyl]benzene TCM tetrakis[4-(carboxyphenyl)-oxamethyl]methane TCMO tetrakis[4-(carboxyphenyl)-oxamethyl]methanoate TCPB 1,2,4,5-tetrakis(4-carboxyphenyl)benzene TCPBDA N,N,N0 ,N0 -tetrakis(4-carboxyphenyl)-biphenyl-4,40 -diamine TDCPTM 4,40 ,400 ,4000 -tetrakis[3,5-di(carboxylate)-1-phenyl]tetraphenyl methane TEA tetraethylammonium TED triethylenediamine tImb 1,3,5-tris(1H-imidazol-4-yl)benzene TMA tetramethylammonium Tp hydrotris(pyrazolyl)borate TPBTM N,N0 ,N00 -tris(isophthalyl)-1,3,5-benzenetricarboxamide 773 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews tpom TP TZI 2,7-ndc 2-stp 3-mepy 4-bcba-H3 4,40 -bpe 4-mIm 5-MeO-ip 5-NO-ip REVIEW tetrakis(4-pyridyloxymethylene)methane 2-tetrazolepyrimidine tetrazolylisophthalate naphthalene-2,7-dicarboxylate 2-sulfonylterephthalate 3-methylpyridine bis(4-carboxy-benzyl)amine trans-bis(4-pyridyl)ethylene 4-methylimidazolate 5-methoxyisophthalate 5-nitro-isophthalate (29) Botzen, W. J. W.; Gowdy, J. M.; Van Den Bergh, J. C. J. M. Clim. Policy 2008, 8, 569. (30) Rochelle, G. T. Science 2009, 325, 1652. (31) Granite, E. J.; Pennline, H. W. Ind. Eng. Chem. Res. 2002, 41, 5470. (32) Lide, D. R. CRC Handbook of Chemistry and Physics; CRC Press: Boca Raton, FL, 2004. (33) Yamposkii, Y.; Pinnau, I.; Freeman, B. D. Materials Science of Membranes for Gas and Vapor Separation; John Wiley & Sons Ltd.: West Sussex, England, 2006. (34) Benson, S. M.; Surles, T. Proc. IEEE 2006, 94, 1795. (35) Aydin, G.; Karakurt, I.; Aydiner, K. Energy Policy 2010, 38, 5072. (36) Khoo, H. H.; Tan, R. B. H. Energy Fuels 2006, 20, 1914. (37) Service, R. F. Science 2009, 325, 1644. (38) Metz, B.; Davidson, O.; de Coninck, H.; Loos, M.; Meyer, L. Special Report on Carbon Dioxide Capture and Storage; Cambridge Univ. Press: Cambridge, U.K., 2006. (39) Klara, S. M.; Srivastava, R. D. Environ. Prog. 2002, 21, 247. (40) da Silva, E. F.; Svendsen, H. F. Int. J. Greenhouse Gas Control 2007, 1, 151. (41) le Bouhelec, E. B.; Mougin, P.; Barreau, A.; Solimando, R. Energy Fuels 2007, 21, 2044. (42) Peeters, A. N. M.; Faaij, A. P. C; Turkenburg, W. C. Int. J. Greenhouse Gas Control 2007, 1, 396. (43) Oyenekan, B. A.; Rochelle, G. T. AIChE J. 2007, 53, 3144. (44) Freeman, S. A.; Dugas, R.; Van Wagener, D. H.; Nguyen, T.; Rochelle, G. T. Int. J. Greenhouse Gas Control 2010, 4, 119. (45) Karadas, F.; Atilhan, M.; Aparicio, S. Energy Fuels 2010, 24, 5817. (46) Idem, R.; Wilson, M.; Tontiwachwithikul, P.; Chakma, A.; Veawab, A.; Aroonwilas, A.; Gelowitz, D. Ind. Eng. Chem. Res. 2006, 45, 3414. (47) Weiland, R. H.; Dingman, J. C.; Cronin, D. B. J. Chem. Eng. Data 1997, 42, 1004. (48) Mason, J. A.; Sumida, K.; Herm, Z. R.; Krishna, R.; Long, J. R. Energy Environ. Sci. 2011, 4, 3030.  ejka, J.; Corma, A.; Zones, S. Zeolites and Catalysis: Synthesis, (49) C Reactions and Applications; Wiley-VCH: Weinheim, Germany, 2010. (50) Kusakabe, K.; Kuroda, T.; Murata, A.; Morooka, S. Ind. Eng. Chem. Res. 1997, 36, 649. (51) Cavenati, S.; Grande, C. A.; Rodrigues, A. E. J. Chem. Eng. Data 2004, 49, 1095. (52) Cavenati, S.; Grande, C. A.; Rodrigues, A. E. Chem. Eng. Sci. 2006, 61, 3893. (53) Himeno, S.; Tomita, T.; Suzuki, K.; Yoshida, S. Microporous Mesoporous Mater. 2007, 98, 62. (54) Ghoufi, A.; Gaberova, L.; Rouquerol, J.; Vincent, D.; Llewellyn, P. L.; Maurin, G. Microporous Mesoporous Mater. 2009, 119, 117. (55) Lee, J.-S.; Kim, J.-H.; Kim, J.-T.; Suh, J.-K.; Lee, J.-M.; Lee, C.-H. J. Chem. Eng. Data 2002, 47, 1237. (56) Wang, Y.; LeVan, M. D. J. Chem. Eng. Data 2009, 54, 2839. (57) O’Keefe, M.; Peskov, M. A.; Ramsden, S. J.; Yaghi, O. M. Acc. Chem. Res. 2008, 41, 1782. (58) Jiang, J.; Yu, J.; Corma, A. Angew. Chem., Int. Ed. 2010, 49, 3120. (59) Palomino, M.; Corma, A.; Rey, F.; Valencia, S. Langmuir 2010, 26, 1910. (60) Zhang, J.; Webley, P. A.; Xiao, P. Energy Convers. Manage. 2008, 49, 346. (61) Li, G.; Xiao, P.; Webley, P.; Zhang, J.; Singh, R.; Marshall, M. Adsorption 2008, 14, 415. (62) Li, G.; Xiao, P.; Webley, P.; Zhang, J.; Singh, R. Energy Procedia 2009, 1, 1123. (63) Konduru, N.; Linder, P.; Assaf-Anid, N. M. AIChE J. 2007, 53, 3137. (64) Jadhav, P. D.; Chatti, R. V.; Biniwale, R. B.; Labhsetwar, N. K.; Devotta, S.; Rayalu, S. S. Energy Fuels 2007, 21, 3555. (65) Zhang, J.; Singh, R.; Webley, P. A. Microporous Mesoporous Mater. 2008, 111, 478. REFERENCES (1) Quadrelli, R.; Peterson, S. Energy Policy 2007, 35, 5938. (2) Pachauri, R. K.; Reisinger, A. IPCC Fourth Assessment Report, Intergovernmental Panel on Climate Change, 2007. (3) U.S. Environmental Protection Agency: http://www.epa.gov/ climatechange/emissions/co2_human.html. (4) Carbon Capture and Storage, Full-Scale Demonstration Progress Update: www.iea.org/G8/docs/ccs_g8july09.pdf, OECD/IEA, 2009. (5) International Energy Agency (IEA): http://www.iea.org/ index_info.asp?id=1959, 2011. (6) Ritter, S. K. Chem. Eng. News 2007, 85, 7. (7) Kumar, B.; Smieja, J. M.; Kubiak, C. P. J. Phys. Chem. C 2010, 114, 14220. (8) Haszeldine, R. S. Science 2009, 325, 1644. (9) Choi, S.; Drese, J. H.; Jones, C. W. ChemSusChem 2009, 2, 796. (10) Li, J.-R.; Kuppler, R. J.; Zhou, H.-C. Chem. Soc. Rev. 2009, 38, 1477. (11) Keskin, S.; van Heest, T. M.; Sholl, D. S. ChemSusChem 2010, 3, 879. (12) D’Alessandro, D. M.; Smit, B.; Long, J. R. Angew. Chem., Int. Ed. 2010, 49, 6058. (13) Ma, S.; Zhou, H.-C. Chem. Commun. 2010, 46, 44. (14) Willis, R. R.; Benin, A. I.; Snurr, R. Q.; Yazaydin, A. O. Nanotechnology for Carbon Dioxide Capture; Wiley-VCH Verlag GmbH & Co.: Weinheim, Germany, 2010. (15) Li, J.-R.; Ma, Y.; McCarthy, M. C.; Sculley, J.; Yu, J.; Jeong, H.-K.; Balbuena, P. B.; Zhou, H.-C. Coord. Chem. Rev. 2011, 255, 1791. (16) Ferey, G.; Serre, C.; Devic, T.; Maurin, G.; Jobic, H.; Llewellyn, P. L.; De Weireld, G.; Vimont, A.; Daturi, M.; Chang, J.-S. Chem. Soc. Rev. 2011, 40, 550. (17) Morris, R. E.; Wheatley, P. S. Angew. Chem., Int. Ed. 2008, 47, 4966. (18) Phan, A.; Doonan, C. J.; Uribe-Romo, F. J.; Knobler, C. B.; O’Keefe, M.; Yaghi, O. M. Acc. Chem. Res. 2010, 43, 58. (19) Simmons, J. M.; Wu, H.; Zhou, W.; Yildirim, T. Energy Environ. Sci. 2011, 4, 2177. (20) Earth System Research Laboratory: http://www.esrl.noaa.gov/ gmd/ccgg/trends/index.html, 2011. (21) Fischer, H.; Wahlen, M.; Smith, J.; Mastroianni, D.; Deck, B. Science 1999, 283, 1712. (22) Demicco, R. V.; Lowenstein, T. K.; Hardie, L. A. Geology 2003, 31, 793. (23) Najam, A.; Rahman, A. A.; Huq, S.; Sokona, Y. Clim. Policy 2003, 3, S9. (24) Demirbas, A. Energy Sources, Part A 2008, 30, 70. (25) Petron, G.; Tans, P.; Frost, G.; Chao, D.; Trainer, M. J. Geophys. Res. 2008, 113, 1. (26) U.S. Energy Information Administration: http://www.eia.gov/ electricity/monthly/index.cfm, 2011. (27) Shafiee, S.; Topal, E. Energy Policy 2009, 37, 181. (28) International Energy Outlook: DOE/EIA-0484, U.S. Energy Information Administration, 2010 http://www.eia.gov/forecasts/ieo/ index.cfm (Accessed March 2010). 774 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW (66) Martin, C. F.; Plaza, M. G.; Pis, J. J.; Rubiera, F.; Pevida, C.; Centeno, T. A. Sep. Purif. Technol. 2010, 74, 225. (67) Silvestre-Albero, J.; Wahby, A.; Sepulveda, E.; Martínez-Escandell, M.; Kaneko, K.; Rodríguez-Reinoso, F. Chem. Commun. 2011, 47, 6840. (68) Plaza, M. G.; García, S.; Rubiera, F.; Pis, J. J.; Pevida, C. Chem. Eng. J. 2010, 163, 41. (69) Plaza, M. G.; Pevida, C.; Arenillas, A.; Rubiera, F.; Pis, J. J. Fuel 2007, 86, 2204. (70) Plaza, M. G.; Pevida, C.; Arias, B.; Casal, M. D.; Martín, C. F.; Fermosa, J.; Rubiera, F.; Pis, J. J. J. Environ. Eng. 2009, 135, 426. (71) Bezerra, D. P.; Oliviera, R. S.; Vieira, R. S.; Cavalcante, C. L., Jr.; Azevedo, D. C. S. Adsorption 2011, 17, 235. (72) El-Kaderi, H. M.; Hunt, J. R.; Medoza-Cortez, J. L.; C^ote, A. P.; Taylor, R. E.; O’Keefe, M.; Yaghi, O. M. Science 2007, 316, 268. (73) C^ote, A. P.; El-Kaderi, H. M.; Furukawa, H.; Hunt, J. R.; Yaghi, O. M. J. Am. Chem. Soc. 2007, 129, 12914. (74) Han, S. S.; Furukawa, H.; Yaghi, O. M.; Goddard, W. A. J. Am. Chem. Soc. 2008, 130, 11580. (75) Glover, T. G.; Hunt, J. R.; Yaghi, O. M. Nature 2010, 2, 235. (76) Dogru, M.; Sonnauer, A.; Gavryushin, A.; Knochel, P.; Bein, T. Chem. Commun. 2011, 47, 1707. (77) Choi, Y. J.; Choi, J. H.; Choi, K. M.; Kang, J. K. J. Mater. Chem. 2011, 21, 1073. (78) Hicks, J. C.; Drese, J. H.; Fauth, D. J.; Gray, M. L.; Qi, G.; Jones, C. W. Microporous Mesoporous Mater. 2003, 62, 29. (79) Xu, X.; Song, C.; Andresen, J. M.; Miller, B. G.; Scaroni, A. W. Microporous Mesoporous Mater. 2003, 62, 29. (80) Jones, J. J. Am. Chem. Soc. 2008, 130, 2902. (81) Jiang, J.-X.; Su, F.; Trewin, A.; Wood, C. D.; Campbell, N. L.; Niu, H.; Dickinson, C.; Ganin, A. Y.; Rosseinsky, M. J.; Khimyak, Y. Z.; Cooper, A. I. Angew. Chem., Int. Ed. 2007, 46, 8574. (82) Cooper, A. I. Adv. Mater. 2009, 21, 1291. (83) Du, N.; Park, H. B.; Robertson, G. P.; Dal-Cin, M. M.; Visser, T.; Scoles, L.; Guiver, M. D. Nat. Mater. 2011, 10, 372. (84) Martin, C. F.; Stockel, E.; Clowes, R. A.; Adams, D. J.; Cooper, A. I.; Pis, J. J.; Rubiera, F.; Pevida, C. J. Mater. Chem. 2011, 21, 5475. (85) Merel, J.; Marc, C.; Francis, M. Ind. Eng. Chem. Res. 2008, 47, 209. (86) Berger, A. H.; Bhown, A. S. Energy Procedia 2011, 4, 562. (87) Herm, Z. R.; Swisher, J. A.; Smit, B.; Krishna, R.; Long, J. R. J. Am. Chem. Soc. 2011, 133, 5664. (88) Ishibashi, M.; Ota, H.; Akutsu, N.; Umeda, S.; Tajika, M.; Izumi, J.; Yasutake, A.; Kabata, T.; Kageyama, Y. Energy Convers. Manage. 1996, 37, 929. (89) Mulgundmath, V.; Tezel, F. H. Adsorption 2010, 16, 587. (90) Rubin, E. S.; Chao, C.; Rao, A. B. Energy Policy 2007, 35, 4444. (91) House, K. Z.; Harvey, C. F.; Aziz, M. J.; Schrag, D. P. Energy Environ. Sci. 2009, 2, 193. (92) Kitagawa, S.; Kitaura, R.; Noro, S.-I. Angew. Chem., Int. Ed. 2004, 43, 2334. (93) Ferey, G. Chem. Soc. Rev. 2008, 37, 191. (94) Lee, J. Y.; Farha, O. K.; Roberts, J.; Scheidt, K. A.; Nguyen, S. T.; Hupp, J. T. Chem. Soc. Rev. 2009, 38, 1450. (95) Murray, L. J.; Dinca, M.; Long, J. R. Chem. Soc. Rev. 2009, 38, 1294. (96) Kuppler, R. J.; Timmons, D. J.; Fang, Q.-R.; Li, J.-R. M.; T., A.; Young, M. D.; Yuan, D.; Zhao, D.; Zhuang, W.; Zhou, H.-C. Coord. Chem. Rev. 2009, 253, 3042. (97) Farrusseng, D.; Aguado, S.; Pinel, C. Angew. Chem., Int. Ed. 2009, 48, 7502. (98) Corma, A.; García, H.; Llabres, i; Xamena, F. X. Chem. Rev. 2010, 110, 4606. (99) Furukawa, H.; Ko, N.; Go, Y. B.; Aratani, N.; Choi, S. B.; Choi, E.; Yazaydin, A. O.; Snurr, R. Q.; O’Keefe, M.; Kim, J.; Yaghi, O. M. Science 2010, 329, 424. (100) Rowsell, J. L. C.; Yaghi, O. M. J. Am. Chem. Soc. 2006, 128, 1304. (101) Furukawa, H.; Kim, J.; Ockwig, N. W.; O’Keefe, M.; Yaghi, O. M. J. Am. Chem. Soc. 2008, 130, 11650. (102) Wang, Z.; Cohen, S. M. Chem. Soc. Rev. 2009, 38, 1315. (103) Tanabe, K. K.; Cohen, S. M. Chem. Soc. Rev. 2011, 40, 498. (104) Czaja, A. U.; Trukhan, N.; M€uller, U. Chem. Soc. Rev. 2009, 38, 1284. (105) Ferey, G.; Mellot-Draznieks, C.; Serre, C.; Millange, F. Acc. Chem. Res. 2005, 38, 217. (106) Kitagawa, S.; Matsuda, R. Coord. Chem. Rev. 2007, 251, 2490. (107) Klinowski, J.; Paz, F. A. A.; Silva, P.; Rocha, J. Dalton Trans. 2011, 40, 321. (108) Jung, D.-W.; Yang, D.-A.; Kim, J.; Kim, J.; Ahn, W.-S. Dalton Trans. 2010, 39, 2883. (109) Cantwell, C. G.; Brown, A. J.; Sholl, D. S.; Nair, S. Cryst. Growth Des. 2011, 11, 4505. (110) Pichon, A.; James, S. L. CrystEngComm 2008, 10, 1839. (111) Klimakow, M.; Klobes, P.; Thuenemann, A. F.; Rademann, K.; Emmerling, F. Chem. Mater. 2010, 22, 5216. (112) Li, M.; Dinca, M. J. Am. Chem. Soc. 2011, 133, 12926. (113) Li, H.; Eddaoudi, M.; O’Keefe, M.; Yaghi, O. M. Nature 1999, 402, 276. (114) Eddaoudi, M.; Li, H.; Yaghi, O. M. J. Am. Chem. Soc. 2000, 122, 1391. (115) Kaye, S. S.; Dailly, A.; Yaghi, O. M.; Long, J. R. J. Am. Chem. Soc. 2007, 129, 14176. (116) Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O’Keefe, M.; Yaghi, O. M. Science 2002, 295, 469. (117) Cavka, J. H.; Jakobsen, S.; Olsbye, U.; Guillou, N.; Lamberti, C.; Bordiga, S.; Lillerud, K. P. J. Am. Chem. Soc. 2008, 130, 13850. (118) Garibay, S. J.; Cohen, S. M. Chem. Commun. 2010, 46, 7700. (119) Serre, C.; Millange, F.; Thouvenot, C.; Nogues, M.; Marsolier, G.; Lou€er, D.; Ferey, G. J. Am. Chem. Soc. 2002, 124, 13519. (120) Biswas, S.; Ahnfeldt, T.; Stock, N. Inorg. Chem. 2011, 50, 9518. (121) Lin, X.; Telepeni, I.; Blake, A. J.; Dailly, A.; Brown, C. M.; Simmons, J. M.; Zoppi, M.; Walker, G. S.; Thomas, M.; Mays, T. J.; Hubberstey, P.; Champness, N. R.; Schr€oder, M. J. Am. Chem. Soc. 2009, 131, 2159. (122) Ferey, G.; Serre, C. Chem. Soc. Rev. 2009, 38, 1380. (123) Horike, S.; Shimomura, S.; Kitagawa, S. Nat. Chem. 2009, 1, 695. (124) Wang, Z.; Cohen, S. M. J. Am. Chem. Soc. 2007, 129, 12368. (125) Wang, Z.; Cohen, S. M. J. Am. Chem. Soc. 2009, 131, 16675. (126) Savonnet, M.; Baser-Bachi, D.; Bats, N.; Perez-Pellitero, J.; Jeanneau, E.; Lecocq, V.; Pine, C.; Farrusseng, D. J. Am. Chem. Soc. 2010, 132, 4518. (127) Bernt, S.; Guillerm, V.; Serre, C.; Stock, N. Chem. Commun. 2011, 47, 2838. (128) Vimont, A.; Goupil, J.-M.; Lavalley, J.-C.; Daturi, M.; Surble, S.; Serre, C.; Millange, F.; Ferey, G.; Audebrand, N. J. Am. Chem. Soc. 2006, 128, 3218. (129) Dinca, M.; Long, J. R. Angew. Chem., Int. Ed. 2008, 47, 6766. (130) Zhou, W.; Wu, H.; Yildirim, T. J. Am. Chem. Soc. 2008, 130, 15268. (131) Dietzel, P. D. C.; Besikiotis, V.; Blom, R. J. Mater. Chem. 2009, 19, 7362. (132) Chen, B.; Xiang, S.; Qian, G. Acc. Chem. Res. 2010, 43, 1115. (133) Chui, S. S.-Y.; Lo, S. M.-F.; Charmant, J; Orpen, A. G.; Williams, I. D. Science 1999, 283, 1148. (134) Murray, L. J.; Dinca, M.; Yano, J.; Chavan, S.; Bordiga, S.; Brown, C. M.; Long, J. R. J. Am. Chem. Soc. 2010, 132, 7856. (135) Xie, L.; Liu, S.; Gao, C.; Cao, R.; Cao, J.; Sun, C.; Su, Z. Inorg. Chem. 2007, 46, 7782. (136) Feldblyum, J. I.; Liu, M.; Gidley, D. W.; Matzger, A. J. J. Am. Chem. Soc. 2011, 133, 18257. (137) Kramer, M.; Schwarz, U.; Kaskel, S. J. Mater. Chem. 2006, 16, 2245. (138) Dinca, M.; Long, J. R. J. Am. Chem. Soc. 2007, 129, 11172. (139) Dinca, M.; Han, W. S.; Liu, Y.; Dailly, A.; Brown, C. M.; Long, J. R. Angew. Chem., Int. Ed. 2007, 46, 1419. 775 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW (172) Moellmer, J.; Moeller, A.; Driesbach, F.; Glaeser, R.; Staudt, R. Microporous Mesoporous Mater. 2011, 138, 140. (173) Hamon, L.; Jolima^itre, E.; Pirngruber, G. D. Ind. Eng. Chem. Res. 2010, 49, 7497. (174) Liang, Z.; Marshall, M.; Chaffee, A. L. Energy Procedia 2009, 1, 1265. (175) Liang, Z.; Marshall, M.; Chaffee, A. L. Energy Fuels 2009, 23, 2785. (176) Liang, Z.; Marshall, M.; Chaffee, A. L. Microporous Mesoporous Mater. 2010, 132, 305. (177) Nune, S. K.; Thallapally, P. K.; Dohnalkova, A.; Wang, C.; Liu, J.; Exarhos, G. J. Chem. Commun. 2010, 46, 4878. (178) Bourrelly, S.; Llewellyn, P. L.; Serre, C.; Millange, F.; Loiseau, T.; Ferey, G. J. Am. Chem. Soc. 2005, 127, 13519. (179) Park, H. J.; Suh, M. P. Chem. Commun. 2010, 46, 610. (180) Senkovska, I.; Hoffmann, F.; Fr€oba, M.; Getzschmann, J.; B€ohlmann, W.; Kaskel, S. Microporous Mesoporous Mater. 2009, 122, 93. (181) Chowdhury, P.; Bikkina, C.; Gumma, S. J. Phys. Chem. C 2009, 113, 6616. (182) Galli, S.; Masciocchi, N.; Tagliabue, G.; Sironi, A.; Navarro, J. A. R.; Salas, J. M.; Mendez-Li~ nan, L.; Domingo, M.; Perez-Mendoza, M.; Barea, E. Chem.—Eur. J. 2008, 14, 9890. (183) Zhang, J.; Xue, Y.-S.; Liang, L.-L.; Ren, S.-B.; Li, Y.-Z.; Du, H.-B.; You, X.-Z. Inorg. Chem. 2010, 49, 7685. (184) Weirsum, A. D.; Soubeyrand-Lenoir, E.; Yang, Q.; Moulin, B.; Guillerm, V.; Yahia, M. B.; Bourrelly, S.; Vimont, A.; Miller, S.; Vagner, C.; Daturi, M.; Guillaume, C.; Serre, C.; Maurin, G.; Llewellyn, P. L. Chem.—Asian J. 2011, DOI: 10.1002/asia.201100201. (185) Rabone, J.; Yue, Y.-F.; Chong, S. Y.; Stylianou, K. C.; Basca, J.; Bradshaw, D.; Darling, G. R.; Berry, N. G.; Khimyak, Y. Z.; Ganin, A. Y.; Wiper, P.; Claridge, J. B.; Rosseinsky, M. J. Science 2010, 329, 1053. (186) Loiseau, T.; Lecroq, L.; Volkringer, C.; Marrot, J.; Ferey, G.; Haouas, M.; Taulelle, F.; Bourrelly, S.; Llewellyn, P. L.; Latroche, M. J. Am. Chem. Soc. 2006, 128, 10223. (187) Surble, S.; Millange, F.; Serre, C.; Duren, T.; Latroche, M.; Bourrelly, S.; Llewellyn, P. L.; Ferey, G. J. Am. Chem. Soc. 2006, 128, 14889. (188) Bao, Z.; Yu, L.; Ren, Q.; Lu, X.; Deng, S. J. Colloid Interface Sci. 2011, 353, 549. € Snurr, R. Q.; Park, T.-H.; Koh, K.; Liu, J.; (189) Yazaydin, A. O.; LeVan, M. D.; Benin, A. I.; Jakubczak, P.; Lanuza, M.; Galloway, D. B.; Low, J. L.; Willis, R. R. J. Am. Chem. Soc. 2009, 131, 18198. (190) Britt, D.; Furukawa, H.; Wang, B.; Glover, T. G.; Yaghi, O. M. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 20637. € Benin, A. I.; Faheem, S. A.; Jakubczak, P.; (191) Yazaydin, A. O.; Low, J. L.; Willis, R. R.; Snurr, R. Q. Chem. Mater. 2009, 21, 1425. (192) Liu, J.; Wang, Y.; Benin, A. I.; Jakubczak, P.; Willis, R. R.; LeVan, M. D. Langmuir 2010, 26, 14301. (193) Dietzel, P. D. C.; Johnsen, R. E.; Fjellvag, H.; Bordiga, S.; Groppo, E.; Chavan, S.; Blom, R. Chem. Commun. 2008, 5125. (194) Aprea, P.; Caputo, D.; Gargiulo, N.; Iucolano, F.; Pepe, F. J. Chem. Eng. Data 2010, 55, 3655. (195) Chowdhury, P.; Bikkina, C.; Meister, D.; Dreisbach, F.; Gumma, S. Microporous Mesoporous Mater. 2009, 117, 406. (196) Farrusseng, D.; Daniel, C.; Gaudillere, C.; Ravon, U.; Schuurman, Y.; Mirodatos, C.; Dubbeldam, D.; Frost, H.; Snurr, R. Q. Langmuir 2009, 25, 7383. (197) Kim, J.; Yang, S.-T.; Choi, S. B.; Sim, J.; Kim, J.; Ahn, W.-S. J. Mater. Chem. 2011, 21, 3070. (198) McDonald, T. M.; D’Alessandro, D. M.; Krishna, R.; Long, J. R. Chem. Sci. 2011, 2, 2022. (199) An, J.; Geib, S. J.; Rosi, N. J. Am. Chem. Soc. 2009, 132, 38. (200) Demessence, A.; D’Alessandro, D. M.; Foo, M. L.; Long, J. R. J. Am. Chem. Soc. 2009, 131, 8784. (201) Vaidhyanathan, R.; Iremonger, S. S.; Dawson, K. W.; Shimizu, G. K. H. Chem. Commun. 2009, 5230. (202) Prasad, T. K.; Hong, D. H.; Suh, M. P. Chem.—Eur. J. 2010, 16, 14043. (140) Sumida, K.; Horike, S.; Kaye, S. S.; Herm, Z. R.; Queen, W. L.; Brown, C. M.; Grandjean, F.; Long, G. J.; Dailly, A.; Long, J. R. Chem. Sci. 2010, 1, 184. (141) Barthelet, K.; Marrot, J.; Riou, D.; Ferey, G. Angew. Chem., Int. Ed. 2002, 41, 281. (142) Ferey, G.; Latroche, M.; Serre, C.; Millange, F.; Loiseau, T.; Percheron-Guegan, A. Chem. Commun. 2003, 2976. (143) Miller, S. R.; Wright, P. A.; Serre, C.; Loiseau, T.; Marrot, J.; Ferey, G. Chem. Commun. 2005, 3850. (144) Whitfield, T. R.; Wang, X.; Liu, L.; Jacobson, A. J. Solid State Sci. 2005, 7, 1096. (145) Anokhina, E. V.; Vougo-Zanda, M.; Wang, X.; Jacobson, A. J. J. Am. Chem. Soc. 2005, 127, 15000.  H.; Kongshaug, K. O.; Swang, O.; Blom, R. (146) Arstad, B.; Fjellvag, Adsorption 2008, 14, 755. (147) Couck, S.; Denayer, J. F. M.; Baron, G. V.; Remy, T.; Gascon, J.; Kapteijn, F. J. Am. Chem. Soc. 2009, 131, 6326. (148) Surble, S.; Serre, C.; Mellot-Draznieks, C.; Millange, F.; Ferey, G. Chem. Commun. 2006, 284. (149) Britt, D.; Tranchemontagne, D.; Yaghi, O. M. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 11623. (150) Caskey, S. R.; Wong-Foy, A. G.; Matzger, A. J. J. Am. Chem. Soc. 2008, 130, 10870. (151) Bauer, S.; Serre, C.; Devic, T.; Horcajada, P.; Marrot, J.; Ferey, G.; Stock, N. Inorg. Chem. 2008, 47, 7568. (152) Biemmi, E.; Christian, S.; Stock, N.; Bein, T. Microporous Mesoporous Mater. 2009, 117, 111. (153) Ahnfeldt, T.; Guillou, N.; Gunzelmann, D.; Margiolaki, I.; Loiseau, T.; Ferey, G.; Senker, J.; Stock, N. Angew. Chem., Int. Ed. 2009, 48, 5163. (154) Palanikumar, M.; Stock, N. Inorg. Chem. 2011, 50, 5085. (155) Low, J. J.; Benin, A. I.; Jakubczak, P.; Abrahamian, J. F.; Faheem, S. A.; Willis, R. R. J. Am. Chem. Soc. 2009, 131, 15834. (156) Tan, J. C.; Cheetham, A. K. Chem. Soc. Rev. 2011, 40, 1059. (157) Chapman, K. W.; Halder, G. J.; Chupas, P. J. J. Am. Chem. Soc. 2008, 130, 10524. (158) Huang, B. L.; Ni, Z.; Millward, A. R.; McGaughey, A. J. H.; Uher, C.; Kaviany, M.; Yaghi, O. M. Mass Transfer 2007, 50, 405. (159) Jakubinek, M. B.; Zhan, B.-Z.; White, M. A. Microporous Mesoporous Mater. 2007, 103, 108. (160) Millward, A. R.; Yaghi, O. M. J. Am. Chem. Soc. 2005, 127, 17998. (161) Farha, O. K.; Yazaydin, A. O.; Eryazici, I.; Malliakas, C. D.; Hauser, B. G.; Kanatzidis, M. G.; Nguyen, S. T.; Snurr, R. Q.; Hupp, J. T. Nat. Chem. 2010, 2, 944. (162) Xue, M.; Liu, Y.; Schaffino, R. M.; Xiang, S.; Zhao, X.; Zhu, G.-S.; Qiu, S.-L.; Chen, B. Inorg. Chem. 2009, 48, 4649. (163) Botas, J. A.; Calleja, G.; Sanchez-Sanchez, M.; Orcajo, M. G. Langmuir 2010, 26, 5300. (164) Choi, J.-S.; Son, W.-J.; Kim, J.; Ahn, W.-S. Microporous Mesoporous Mater. 2008, 116, 727. (165) Gedrich, K.; Senkovska, I.; Klein, N.; Stoeck, U.; Henshcel, A.; Lohe, M. R.; Baburin, I. A.; Mueller, U.; Kaskel, S. Angew. Chem., Int. Ed. 2010, 49, 8489. (166) Yuan, D.; Zhao, D.; Sun, D.; Zhou, H.-C. Angew. Chem., Int. Ed. 2010, 49, 5357. (167) Llewellyn, P. L.; Bourrelly, S.; Serre, C.; Vimont, A.; Daturi, M.; Hamon, L.; Weireld, G. D.; Chang, J.-S.; Hong, D.-Y.; Hwang, Y. K.; Jhung, S. H.; Ferey, G. Langmuir 2008, 24, 7245. (168) Zhang, Z.; Huang, S.; Xian, S.; Xi, H.; Li, Z. Energy Fuels 2011, 25, 835. (169) Mu, B.; Schoenecker, P. M.; Walton, K. S. J. Phys. Chem. C 2010, 114, 6464. (170) Tan, C.; Yang, S.; Champness, N. R.; Lin, X.; Blake, A. J.; Lewis, W.; Schr€oder, M. Chem. Commun. 2011, 47, 4487. (171) Park, Y. K.; Choi, S. B.; Kim, H.; Kim, K.; Won, B.-H.; Choi, K.; Choi, J.-S.; Ahn, W.-S.; Won, N.; Kim, S.; Jung, D. H.; Choi, S.-H.; Kim, G.-H.; Cha, S.-S.; Jhon, Y. H.; Yang, J. K.; Kim, J. Angew. Chem., Int. Ed. 2007, 46, 8230. 776 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW (235) Pachfule, P.; Das, R.; Poddar, P.; Banerjee, R. Cryst. Growth Des. 2011, 11, 1215. (236) Zhang, J.; Wu, H.; Emge, T. J.; Li, J. Chem. Commun. 2010, 46, 9152. (237) Thetiot, F.; Duhayon, C.; Venkatakrishnan, T. S.; Sutter, J.-P. Cryst. Growth Des. 2008, 8, 1870. (238) Thallapally, P. K.; Tian, J.; Kishan, M. R.; Fernandez, C. A.; Dalgarno, S. J.; McGrail, P. B.; Warren, J. E.; Atwood, J. L. J. Am. Chem. Soc. 2008, 130, 16842. (239) Kishan, M. R.; Tian, J.; Thallapally, P. K.; Fernandez, C. A.; Dalgarno, S. J.; Warren, J. E.; McGrail, B. P.; Atwood, J. L. Chem. Commun. 2010, 46, 538. (240) Pachfule, P.; Das, R.; Poddar, P.; Banerjee, R. Cryst. Growth Des. 2010, 10, 2475. (241) Park, H. J.; Cheon, Y. E.; Suh, M. P. Chem.—Eur. J. 2010, 16, 11662. (242) Chen, S.-S.; Chen, M.; Takamizawa, S.; Chen, M.-S.; Su, Z.; Sun, W.-Y. Chem. Commun. 2011, 47, 752. (243) Wu, H.; Reali, R. S.; Smith, D. A.; Trachtenberg, M. C.; Li, J. Chem.—Eur. J. 2010, 16, 13951. (244) Fernandez, C. A.; Thallapally, P. K.; Motkuri, R. K.; Nune, S. K.; Sumrak, J. C.; Tian, J.; Liu, J. Cryst. Growth Des. 2010, 10, 1037. (245) Farha, O. K.; Spokoyny, A. M.; Mulfort, K. L.; Hawthorne, M. F.; Mirkin, C. A.; Hupp, J. T. J. Am. Chem. Soc. 2007, 129, 12680. (246) Bae, Y.-S.; Farha, O. K.; Spokoyny, A. M.; Mirkin, C. A.; Hupp, J. T.; Snurr, R. Q. Chem. Commun. 2008, 4135. (247) Wang, B.; C^ote, A. P.; Furukawa, H.; O’Keefe, M.; Yaghi, O. M. Nature 2008, 453, 207. (248) Zhang, J.-P.; Chen, X.-M. J. Am. Chem. Soc. 2009, 131, 5516. (249) Pachfule, P.; Panda, T.; Dey, C.; Banerjee, R. CrystEngComm 2010, 12, 2381. (250) Miller, S. R.; Wright, P. A.; Devic, T.; Serre, C.; Ferey, G.; Llewellyn, P. L.; Denoyel, R.; Gaberova, L.; Filinchuk, Y. Langmuir 2009, 25, 3618. (251) Zhang, Y.-J.; Liu, T.; Kanegawa, S.; Sato, O. J. Am. Chem. Soc. 2010, 132, 912. (252) Mallick, A.; Saha, S.; Pachfule, P.; Roy, S.; Banerjee, R. J. Mater. Chem. 2010, 20, 9073. (253) Bao, Z.; Alnemrat, S.; Yu, L.; Vasiliev, I.; Ren, Q.; Lu, X.; Deng, S. J. Colloid Interface Sci. 2011, 357, 504. (254) Bae, Y.-S.; Spokoyny, A. M.; Farha, O. K.; Snurr, R. Q.; Hupp, J. T.; Mirkin, C. A. Chem. Commun. 2010, 46, 3478. (255) Gandara, F.; Gutierrez-Puebla, E.; Iglesias, M.; Snejko, N.; Monge, M. A. Cryst. Growth Des. 2010, 10, 128. (256) Culp, J. T.; Goodman, A. L.; Chirdon, D.; Sankar, S. G.; Matranga, C. J. Phys. Chem. C 2010, 114, 2184. (257) Lambert, J. B.; Liu, Z.; Liu, C. Organometallics 2008, 27, 1464. (258) Lee, Y.-G.; Moon, H. R.; Cheon, Y. E.; Suh, M. P. Angew. Chem., Int. Ed. 2008, 47, 7741. (259) Guo, X.; Zhu, G.; Li, Z.; Sun, F.; Yang, Z.; Qiu, S. Chem. Commun. 2006, 3172. (260) Hu, Y.; Xiang, S.; Zhang, W.; Zhang, Z.; Wang, L.; Bai, J.; Chen, B. Chem. Commun. 2009, 7551. (261) Si, X.; Jiao, C.; Li, F.; Zhang, J.; Wang, S.; Liu, S.; Li, Z.; Sun, L.; Xu, F.; Gabelica, Z.; Schick, C. Energy Environ. Sci. 2011, 4, 4522. (262) Bae, Y.-S.; Farha, O. K.; Hupp, J. T.; Snurr, R. Q. J. Mater. Chem. 2009, 19, 2131. (263) Moellmer, J.; Celer, E. B.; Luebke, R.; Cairns, A. J.; Staudt, R.; Eddaoudi, M.; Thommes, M. Microporous Mesoporous Mater. 2010, 129, 345. (264) Farha, O. K.; Spokoyny, A. M.; Mulfort, K. L.; Galli, S.; Hupp, J. T.; Mirkin, C. A. Small 2009, 15, 1727. (265) Kondo, A.; Chinen, A.; Kajiro, H.; Nakagawa, T.; Kato, K.; Takata, M.; Hattori, Y.; Okino, F.; Ohba, T.; Kaneko, K.; Kanoh, H. Chem.—Eur. J. 2009, 15, 7549. (266) Cheng, Y.; Kondo, A.; Noguchi, H.; Kajiro, H.; Urita, K.; Ohba, T.; Kaneko, K.; Kanoh, H. Langmuir 2009, 25, 4510. (267) Wu, S.; Ma, L.; Long, L.-S.; Zheng, L.-S.; Lin, W. Inorg. Chem. 2009, 48, 2436. (203) Kanoh, H.; Kondo, A.; Noguchi, H.; Kajiro, H.; Tohdoh, A.; Hattori, Y.; Xu, W.-C.; Inoue, M.; Sugiura, T.; Morita, K.; Tanaka, H.; Ohba, T.; Kaneko, K. J. Colloid Interface Sci. 2009, 334, 1. (204) Bloch, E. D.; Britt, D.; Lee, C.; Doonan, C. J.; Uribe-Romo, F. J.; Furukawa, H.; Long, J. R.; Yaghi, O. M. J. Am. Chem. Soc. 2010, 132, 14382. (205) Chen, S.-S.; Chen, M.; Takamizawa, S.; Wang, P.; Lv, G.-C.; Sun, W.-Y. Chem. Commun. 2011, 47, 4902. (206) Kim, T. K.; Suh, M. P. Chem. Commun. 2011, 47, 4258. (207) Navarro, J. A. R.; Barea, E.; Salas, J. M.; Masciocchi, N.; Galli, S.; Sironi, A.; Ania, C. O.; Parra, J. B. J. Mater. Chem. 2007, 17, 1939. (208) Rallapalli, P.; Prasanth, K. P.; Patil, D.; Somani, R. S.; Jasra, R. V.; Bajaj, H. C. J. Porous Mater. 2011, 18, 205. (209) Miller, S. R.; Pearce, G. M.; Wright, P. A.; Bonino, F.; Chavan, S.; Bordiga, S.; Margiolaki, I.; Guillou, N.; Ferey, G.; Bourrelly, S.; Llewellyn, P. L. J. Am. Chem. Soc. 2008, 130, 15967. (210) Stylianou, K. C.; Warren, J. E.; Chong, S. Y.; Rabone, J.; Basca, J.; Bradshaw, D.; Rosseinsky, M. J. Chem. Commun. 2011, 47, 3389. (211) Chen, Z.; Xiang, S.; Arman, H. D.; Li, P.; Tidrow, S.; Zhao, D.; Chen, B. Eur. J. Inorg. Chem. 2010, 24, 3745. (212) Southon, P. D.; Liu, L.; Fellows, E. A.; Price, D. J.; Halder, G. J.; Chapman, K. W.; Moubaraki, B.; Murray, K. S.; Letard, J.-F.; Kepert, C. J. J. Am. Chem. Soc. 2009, 131, 10998. (213) Banerjee, R.; Furukawa, H.; Britt, D.; Knobler, C.; O’Keefe, M.; Yaghi, O. M. J. Am. Chem. Soc. 2009, 131, 3875. (214) Morris, W.; Leung, B.; Furukawa, H.; Yaghi, O. K.; He, N.; Hayashi, H.; Houndonougbo, Y.; Asta, M.; Laird, B. B.; Yaghi, O. M. J. Am. Chem. Soc. 2010, 132, 11006. (215) Debatin, F.; Thomas, A.; Kelling, A.; Hedin, N.; Bacsik, Z.; Senkovska, I.; Kaskel, S.; Junginger, M.; Muller, H.; Schilde, U.; Jager, C.; Friedrich, A.; Holdt, H.-J. Angew. Chem., Int. Ed. 2010, 49, 1258. (216) Zhao, Z.; Li, Z.; Lin, Y. S. Ind. Eng. Chem. Res. 2009, 48, 10015. (217) Saha, D.; Bao, Z.; Jia, F.; Deng, S. Environ. Sci. Technol. 2010, 44, 1820. (218) Lu, C.-M.; Liu, J.; Xiao, K.; Harris, A. T. Chem. Eng. J. 2010, 156, 465. (219) Deng, H.; Doonan, C. J.; Furukawa, H.; Ferriera, R. B.; Towne, J.; Knobler, C. B.; Wang, B.; Yaghi, O. M. Science 2010, 327, 846. (220) Llewellyn, P. L.; Bourrelly, S.; Serre, C.; Filinchuk, Y.; Ferey, G. Angew. Chem., Int. Ed. 2006, 45, 7751. (221) Chen, Z.; Liu, X.; Zhang, C.; Zhang, Z.; Liang, F. Dalton Trans. 2010, 40, 1911. (222) Chen, Z.; Xiang, S.; Zhao, D.; Chen, B. Cryst. Growth Des. 2009, 9, 5293. (223) Zhang, Z.; Xiang, S.; Rao, X.; Zheng, Q.; Fronczek, F. R.; Qian, G.; Chen, B. Chem. Commun. 2010, 46, 7205. (224) Beobide, G.; Wang, W.-G.; Castillo, O.; Luque, A.; Roman, P.; Tagliabue, G.; Galli, S.; Navarro, J. A. R. Inorg. Chem. 2008, 47, 5267. (225) Navarro, J. A. R.; Barea, E.; Salas, J. M.; Masciocchi, N.; Galli, S.; Sironi, A.; Ania, C. O.; Parra, J. B. Inorg. Chem. 2006, 45, 2397. (226) Banerjee, R.; Phan, A.; Wang, B.; Knobler, C.; Furukawa, H.; O’Keefe, M.; Yaghi, O. M. Science 2008, 319, 939. (227) An, J.; Rosi, N. L. J. Am. Chem. Soc. 2010, 132, 5578. (228) García-Ricard, O. J.; Hernandez-Maldonado, A. J. J. Phys. Chem. C 2010, 114, 1827. (229) Mulfort, K. L.; Farha, O. K.; Malliakas, C. D.; Kanatzidis, M. G.; Hupp, J. T. Chem.—Eur. J. 2010, 16, 276. (230) Kitaura, R.; Matsuda, R.; Kubota, Y.; Kitagawa, S.; Takata, M.; Kobayashi, T. C.; Suzuki, M. J. Phys. Chem. B 2005, 109, 23378. (231) Bae, Y.-S.; Hauser, B. G.; Farha, O. K.; Hupp, J. T.; Snurr, R. Q. Microporous Mesoporous Mater. 2011, 141, 231. (232) Lincke, J.; L€assig, D.; Moellmer, J.; Reichenbach, C.; Puls, A.; Moeller, A.; Gl€aser, R.; Kalies, G.; Staudt, R.; Krautscheid, H. Microporous Mesoporous Mater. 2011, 142, 62. (233) Tian, J.; Motkuri, R. K.; Thallapally, P. K.; McGrail, B. P. Cryst. Growth Des 2010, 10, 5327. (234) Bae, Y.-S.; Dubbeldam, D.; Nelson, A.; Walton, K. S.; Hupp, J. T.; Snurr, R. Q. Chem. Mater. 2009, 21, 4768. 777 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW (304) Bastin, L.; Barcia, P. S.; Hurtado, E. J.; Silva, J. A. C.; Rodrigues, A. E.; Chen, B. J. Phys. Chem. C 2008, 112, 1575. (305) Dinca, M.; Long, J. R. J. Am. Chem. Soc. 2005, 127, 9376. (306) Bae, Y.-S.; Lee, C.-H. Carbon 2005, 43, 95. (307) Myers, A. L.; Prausnitz, J. M. AIChE J. 1965, 11, 121. (308) Keskin, S. J. Phys. Chem. C 2011, 115, 800. (309) Coudert, F.-X.; Mellot-Draznieks, C.; Fuchs, A. J.; Boutin, A. J. Am. Chem. Soc. 2009, 131, 11329. (310) Coudert, F.-X. Phys. Chem. Chem. Phys. 2010, 12, 10904. (311) Zheng, B.; Bai, J.; Duan, J.; Wojtas, L.; Zaworotko, M. J. J. Am. Chem. Soc. 2011, 133, 748. (312) Keskin, S. Ind. Eng. Chem. Res. 2011, 50, 8230. (313) Keskin, S.; Sholl, D. S. Ind. Eng. Chem. Res. 2009, 48, 914. (314) Liu, Y.; Liu, D.; Yang, Q.; Zhong, C.; Mi, J. Ind. Eng. Chem. Res. 2010, 49, 2902. (315) Chen, Y. F.; Jiang, J. W. ChemSusChem 2010, 3, 982. (316) Yang, Q.; Zhong, C. J. Phys. Chem. B 2006, 110, 17776. (317) Krishna, R.; van Baten, J. M. Phys. Chem. Chem. Phys. 2011, 13, 10593. (318) Keskin, S.; Liu, J.; Johnson, J. K.; Sholl, D. S. Langmuir 2008, 24, 8254. (319) Keskin, S. Ind. Eng. Chem. Res. 2010, 49, 11689. (320) Liu, B.; Yang, Q.; Xue, C.; Zhong, C.; Chen, B.; Smit, B. J. Phys. Chem. C 2008, 112, 9854. (321) Liu, B.; Smit, B. Langmuir 2009, 25, 5918. (322) Yang, Q.; Xue, C.; Zhong, C.; Chen, J.-F. AIChE J. 2007, 53, 2832. (323) Richter, E.; Sch€utz, W.; Myers, A. L. Chem. Eng. Sci. 1989, 44, 1609. (324) Chen, H.; Sholl, D. S. Langmuir 2007, 23, 6431. (325) Keskin, S.; Liu, J.; Rankin, R. B.; Johnson, J. K.; Sholl, D. S. Ind. Eng. Chem. Res. 2009, 48, 2355. (326) Kizzie, A. C.; Wong-Foy, A. G.; Matzger, A. J. Langmuir 2011, 27, 6368. (327) Hamon, L.; Llewellyn, P. L.; Devic, T.; Ghoufi, A.; Clet, G.; Guillerm, V.; Pirngruber, G. D.; Maurin, G.; Serre, C.; Driver, G.; van Beek, W.; Jolima^itre, E.; Vimont, A.; Daturi, M.; Ferey, G. J. Am. Chem. Soc. 2009, 131, 17490. (328) Finsy, V.; Ma, L.; Alaerts, L.; De Vos, D. E.; Baron, G. V.; Denayer, J. F. M. Microporous Mesoporous Mater. 2009, 120, 221. (329) Barea, E.; Tagliabue, G.; Wang, W.-G.; Perez-Mendoza, M.; Mendez-Li~ nan, L.; Lopez-Garzon, F. J.; Galli, S.; Masciocchi, N.; Navarro, J. A. R. Chem.—Eur. J. 2010, 16, 931. (330) Nakagawa, K.; Tanaka, D.; Horike, S.; Shimomura, S.; Higuchi, M.; Kitagawa, S. Chem. Commun. 2010, 46, 4258. (331) Warren, J. E.; Pritchard, R. G.; Abram, D.; Davies, H. M.; Savarese, T. L.; Cash, R. J.; Raithby, P. R.; Morris, R.; Jones, R. H.; Teat, S. J. J. Appl. Crystallogr. 2009, 42, 457. (332) Xie, L.-H.; Lin, J.-B.; Liu, X.-M.; Wang, Y.; Zhang, W.-X.; Zhang, J.-P.; Chen, X.-M. Inorg. Chem. 2010, 49, 1158. (333) Lin, J.-B.; Xue, W.; Zhang, J.-P.; Chen, X.-M. Chem. Commun. 2011, 47, 926. (334) Serre, C.; Bourrelly, S.; Vimont, A.; Ramsahye, N. A.; Maurin, G.; Llewellyn, P. L.; Daturi, M.; Filinchuk, Y.; Leynaud, O.; Barnes, P.; Ferey, G. Adv. Mater. 2007, 19, 2246. (335) Loiseau, T.; Serre, C.; Huguenard, C.; Fink, G.; Taulelle, F.; Henry, M.; Bataille, T.; Ferey, G. Chem.—Eur. J. 2004, 10, 1373. (336) Vaidhyanathan, R.; Iremonger, S. S.; Shimizu, G. K. H.; Boyd, P. G.; Alavi, S.; Woo, T. K. Science 2010, 330, 650. (337) Bordiga, S.; Regli, L.; Bonino, F.; Groppo, E.; Lamberti, C.; Xiao, B.; Wheatley, P. S.; Morris, R. E.; Zecchina, A. Phys. Chem. Chem. Phys. 2007, 9, 2676. (338) Vimont, A.; Travert, A.; Bazin, P.; Lavalley, J.-C.; Daturi, M.; Serre, C.; Ferey, G.; Bourrelly, S.; Llewellyn, P. L. Chem. Commun. 2007, 31, 3291. (339) Valenzano, L.; Civalleri, B.; Chavan, S.; Palomino, G. T.; Arean, C. O.; Bordiga, S. J. Phys. Chem. C 2010, 114, 11185. (340) Zhao, Y.; Wu, H.; Emge, T. J.; Gong, Q.; Nijem, N.; Chabal, Y. J.; Kong, L.; Langreth, D. C.; Liu, H.; Zeng, H.; Li, J. Chem.—Eur. J. 2011, 17, 5101. (268) Hayashi, H.; C^ote, A. P.; Furukawa, H.; O’Keefe, M.; Yaghi, O. M. Nat. Mater. 2007, 6, 501. (269) Chandler, B. D.; Cramb, D. T.; Shimizu, G. K. H. J. Am. Chem. Soc. 2006, 128, 10403. (270) Wu, C.-D.; Lin, W. Angew. Chem., Int. Ed. 2005, 44, 1958. (271) Spokoyny, A. M.; Farha, O. K.; Mulfort, K. L.; Hupp, J. T.; Mirkin, C. A. Inorg. Chim. Acta 2010, 364, 266. (272) Gadzikwa, T.; Farha, O. K.; Mulfort, K. L.; Hupp, J. T.; Nguyen, S. T. Chem. Commun. 2009, 3720. (273) Inubushi, Y.; Horike, S.; Fukushima, T.; Akiyama, G.; Matsuda, R.; Kitagawa, S. Chem. Commun. 2010, 46, 9229. (274) Fletcher, A. J.; Cussen, E. J.; Bradshaw, D.; Rosseinsky, M. J.; Thomas, K. M. J. Am. Chem. Soc. 2004, 126, 9750. (275) Park, H. J.; Suh, M. P. Chem.—Eur. J. 2008, 14, 8812. (276) Tian, Y.-Q.; Yao, S.-Y.; Gu, D.; Cui, K.-H.; Guo, D.-W.; Zhang, G.; Chen, Z.-X.; Zhao, D.-Y. Chem.—Eur. J. 2010, 16, 1137. (277) Zheng, S.-T.; Li, Y.; Wu, T.; Nieto, R. A.; Feng, P.; Bu, X. Chem.—Eur. J. 2010, 16, 13035. (278) Wu, C.-D.; Lin, W. Dalton Trans. 2006, 4563. (279) Neofotistou, E.; Malliakas, C. D.; Trikalitis, P. N. Chem.—Eur. J. 2009, 15, 4523. (280) Tan, Y.-X.; Wang, F.; Kang, Y.; Zhang, J. Chem. Commun. 2011, 47, 770. (281) Cheon, Y. E.; Suh, M. P. Chem. Commun. 2009, 2296. (282) Chun, H.; Seo, J. Inorg. Chem. 2009, 48, 9980. (283) Wu, T.; Zhang, J.; Zhou, C.; Wang, L.; Bu, X.; Feng, P. J. Am. Chem. Soc. 2009, 131, 6111. (284) Moon, H. R.; Kobayashi, N.; Suh, M. P. Inorg. Chem. 2006, 45, 8672. (285) Walton, K. S.; Millward, A. R.; Dubbeldam, D.; Frost, H.; Low, J. L.; Yaghi, O. K.; Snurr, R. Q. J. Am. Chem. Soc. 2008, 130, 406. (286) Lama, P.; Aijaz, A.; Neogi, S.; Barbour, L. J.; Bharadwaj, P. K. Cryst. Growth Des. 2010, 10, 3410. (287) Wu, C.-D.; Lin, W. Chem. Commun. 2005, 3673. (288) Chandler, B. D.; Yu, J. O.; Cramb, D. T.; Shimizu, G. K. H. Chem. Mater. 2007, 19, 4467. (289) Pan, L.; Adams, K. M.; Hernandez, H. E.; Wang, X.; Zheng, C.; Hattori, Y.; Kaneko, K. J. Am. Chem. Soc. 2003, 125, 3062. (290) Vaidhyanathan, R.; Liang, J.; Iremonger, S. S.; Shimizu, G. K. H. Supramol. Chem. 2011, 23, 278. (291) Alsobrook, A. N.; Hauser, B. G.; Hupp, J. T.; Alekseev, E. V.; Depmeier, W.; Albrecht-Schmitt, T. E. Chem. Commun. 2010, 46, 9167. (292) Liang, J.; Shimizu, G. K. H. Inorg. Chem. 2007, 46, 10449. (293) Pan, H.; Ritter, J. A.; Balbuena, P. B. Langmuir 1998, 14, 6323. (294) Czepirsky, L.; Jagiello, J. Chem. Eng. Sci. 1989, 44, 797. (295) Gu, J.-M.; Kwon, T.-H.; Park, J.-H.; Huh, S. Dalton Trans. 2010, 39, 5608. (296) Volkringer, C.; Loiseau, T.; Haouas, M.; Taulelle, F.; Popov, D.; Burghammer, M.; Riekel, C.; Zlotea, C.; Cuevas, F.; Latroche, M.; Phanon, D.; Knofelv, C.; Llewellyn, P. L.; Ferey, G. Chem. Mater. 2009, 21, 5783. (297) Xiang, Z.; Hu, Z.; Cao, D.; Yang, W.; Jianmin, L.; Han, B.; Wang, W. Angew. Chem., Int. Ed. 2011, 50, 491. (298) Wang, Q. M.; Shen, D.; B€ulow, M.; Lau, M. L.; Deng, S.; Fitch, F. R.; Lemcoff, N. O.; Semanscin, J. Microporous Mesoporous Mater. 2002, 55, 217. (299) Choi, S. B.; Seo, M. J.; Cho, M.; Kim, Y.; Jin, M. K.; Jung, D.-Y.; Choi, J.-S.; Ahn, W.-S.; Rowsell, J. L. C.; Kim, J. Cryst. Growth Des. 2007, 7, 2290. (300) Lin, J.-B.; Zhang, J.-P.; Chen, X.-M. J. Am. Chem. Soc. 2010, 132, 6654. (301) Culp, J. T.; Smith, M. R.; Bittner, E.; Bockrath, B. J. Am. Chem. Soc. 2008, 130, 12427. (302) Procopio, E. Q.; Linares, F.; Montoro, C.; Colombo, V.; Maspero, A.; Barea, E.; Navarro, J. A. R. Angew. Chem., Int. Ed. 2010, 49, 7308. (303) Barcia, P. S.; Bastin, L.; Hurtado, E. J.; Silva, J. A. C.; Rodrigues, A. E.; Chen, B. Sep. Sci. Technol. 2008, 43, 3494. 778 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW (378) Horike, S.; Dinca, M.; Tamaki, K.; Long, J. R. J. Am. Chem. Soc. 2008, 130, 5854. (379) Vishnayakov, A.; Ravikovitch, P. I.; Neimark, A. V.; B€ulow, M.; Wang, Q. M. Nano Lett. 2003, 3, 713.  (380) Dietzel, P. D. C.; Morita, Y.; Blom, R.; Fjellvag, H. Angew. Chem., Int. Ed. 2005, 44, 6354. (381) Rosi, N. L.; Kim, J.; Eddaoudi, M.; Chen, B.; O’Keefe, M.; Yaghi, O. M. J. Am. Chem. Soc. 2005, 127, 1504.  (382) Dietzel, P. D. C.; Panella, B.; Hirscher, M.; Blom, R.; Fjellvag, H. Chem. Commun. 2006, 959.  (383) Dietzel, P. D. C.; Blom, R.; Fjellvag, H. Eur. J. Inorg. Chem. 2008, 3624. (384) Sumida, K.; Brown, C. M.; Herm, Z. R.; Chavan, S.; Bordiga, S.; Long, J. R. Chem. Commun. 2011, 47, 1157. (385) Huang, L.; Wang, H.; Chen, J.; Wang, Z.; Sun, J.; Zhao, D.; Yan, Y. Microporous Mesoporous Mater. 2003, 58, 105. (386) Schr€ock, K.; Schr€oder, F.; Heyden, M.; Fischer, R. A.; Havenith, M. Phys. Chem. Chem. Phys. 2008, 10, 4732. (387) Hausdorf, S.; Wagler, J.; Mossig, R.; Mertens, F. O. R. L. J. Phys. Chem. A 2008, 112, 7567. (388) Greathouse, J. A.; Allendorf, M. D. J. Am. Chem. Soc. 2006, 128, 10678. (389) Cychosz, K. A.; Matzger, A. J. Langmuir 2010, 26, 17198. (390) Li, Y.; Yang, R. T. Langmuir 2007, 23, 12937. (391) Park, K. S.; Ni, Z.; C^ote, A. P.; Choi, J. Y.; Huang, R.; UribeRomo, F. J.; Chae, H. K.; O’Keefe, M.; Yaghi, O. M. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 10186. (392) Huang, X.-C.; Lin, Y.-Y.; Zhang, J.-P.; Chen, X.-M. Angew. Chem., Int. Ed. 2006, 45, 1557. (393) Gu, J.-Z.; Lu, W.-G.; Jiang, L.; Zhou, H.-C.; Lu, T.-B. Inorg. Chem. 2007, 46, 5835. (394) Biswas, S.; Gryzwa, M.; Nayek, H. P.; Dehnen, S.; Senkovska, I.; Kaskel, S.; Volkmer, D. Dalton Trans. 2009, 6487. (395) Choi, H. J.; Dinca, M.; Dailly, A.; Long, J. R. Energy Environ. Sci. 2010, 3, 117. (396) Galli, S.; Masciocchi, N.; Colombo, V.; Maspero, A.; Palmisiano, G.; Lopez-Garzon, F. J.; Domingo-Garcia, M.; Fernandez-Morales, I.; Barea, E.; Navarro, J. A. R. Chem. Mater. 2010, 22, 1664. (397) Colombo, V.; Galli, S.; Choi, H.-J.; Han, G. D.; Maspero, A.; Palmisiano, G.; Masciocchi, N.; Long, J. R. Chem. Sci. 2011, 2, 1311. (398) Catalan, J.; Abboud, J. L. M.; Elguero, J. Adv. Heterocycl. Chem. 1987, 41, 187. (399) Ebert, C.; Elguero, J.; Musumarra, G. J. Phys. Org. Chem. 1990, 3, 651. (400) Ferey, G.; Serre, C.; Mellot-Draznieks, C.; Millange, F.; Surble, S.; Dutour, J.; Margiolaki, I. Angew. Chem., Int. Ed. 2004, 43, 6296. (401) Ferey, G.; Mellot-Draznieks, C.; Serre, C.; Millange, F.; Dutour, J.; Surble, S.; Margiolaki, I. Science 2005, 309, 2040. (402) Volkringer, C.; Popov, D.; Loiseau, T.; Guillou, N.; Ferey, G.; Haouas, M.; Taulelle, F.; Mellot-Draznieks, C.; Burghammer, M.; Riekel, C. Nat. Mater. 2007, 6, 760. (403) K€usgens, P.; Rose, M.; Senkovska, I.; Fr€ode, H.; Henshcel, A.; Siegle, S.; Kaskel, S. Microporous Mesoporous Mater. 2009, 120, 325. (404) Ehrenmann, J.; Henninger, S. K.; Janiak, C. Eur. J. Inorg. Chem. 2011, 471. (405) Lin, X.; Blake, A. J.; Wilson, C.; Sun, X. Z.; Champness, N. R.; George, M.; Hubberstey, P.; Mokoya, R.; Schr€oder, M. J. Am. Chem. Soc. 2006, 128, 10745. (406) Pan, L.; Parker, B.; Huang, X.; Olson, D. H.; Lee, J. Y.; Li, J. J. Am. Chem. Soc. 2006, 128, 4180. (407) Lee, J. Y.; Olson, D. H.; Pan, L.; Emge, T. J.; Li, J. Adv. Funct. Mater. 2007, 17, 1255. (408) Chen, Y. F.; Lee, J. Y.; Babarao, R.; Li, J.; Jiang, J. W. J. Phys. Chem. C 2010, 114, 6602. (409) Wu, T.; Shen, L.; Luebbers, M.; Hu, C.; Chen, Q.; Ni, Z.; Masel, R. I. Chem. Commun. 2010, 46, 6120. (410) Nguyen, J. G.; Cohen, S. M. J. Am. Chem. Soc. 2010, 132, 4560. (341) Kauffman, K. L.; Culp, J. T.; Goodman, A.; Matranga, C. J. Phys. Chem. C 2011, 115, 1857. (342) Taylor, J. H.; Benedict, W. S.; Strong, J. J. Chem. Phys. 1952, 20, 1884. (343) Yoon, J. W.; Seo, Y.-K.; Hwang, Y. K.; Chang, J.-S.; Leclerc, H.; Wuttke, S.; Bazin, P.; Vimont, A.; Daturi, M.; Bloch, E.; Llewellyn, P. L.; Serre, C.; Horcajada, P.; Greneche, J.-M.; Rodrigues, A. E.; Ferey, G. Angew. Chem., Int. Ed. 2010, 49, 5949. (344) Windisch, C. F., Jr.; Thallapally, P. K.; McGrail, B. P. Spectrochim. Acta A 2009, 74, 629. (345) Gascon, J.; Aktay, U.; Hernandez-Alonso, M. D.; van Klink, G. P. M; Kapteijn, F. J. Catal. 2009, 261, 75. (346) Stavitski, E.; Pidko, E. A.; Couck, S.; Remy, T.; Hensen, E. J. M.; Weckhuysen, B. M.; Denayer, J.; Gascon, J.; Kapteijn, F. Langmuir 2011, 27, 3970. (347) Dubbeldam, D.; Frost, H.; Walton, K. S.; Snurr, R. Q. Fluid Phase Equilib. 2007, 261, 152. (348) Torrisi, A.; Bell, R. G.; Mellot-Draznieks, C. Cryst. Growth Des. 2010, 10, 2839. (349) Babarao, R.; Jiang, J.; Sandler, S. I. Langmuir 2010, 25, 5239. (350) Wu, D.; Xu, Q.; Liu, D.; Zhong, C. J. Phys. Chem. C 2010, 114, 16611. (351) Wells, B. A.; Chaffee, A. L. Adsorption 2011, 17, 255. (352) Walton, K. S.; Millward, A. R.; Dubbeldam, D.; Frost, H.; Low, J. J.; Yaghi, O. M.; Snurr, R. Q. J. Am. Chem. Soc. 2008, 130. (353) Ghoufi, A.; Maurin, G. J. Phys. Chem. C 2010, 114, 6496. (354) Liu, B.; Smit, B. J. Phys. Chem. C 2010, 114, 8515. (355) Valenzano, L.; Civalleri, B.; Chavan, S.; Bordiga, S.; Nilsen, M. H.; Jakobsen, S.; Lillerud, K. P.; Lamberti, C. Chem. Mater. 2011, 23, 1700. (356) Grajciar, L.; Wiersum, A. D.; Llewellyn, P. L.; Chang, J.-S.; Nachtigall, P. J. Phys. Chem. C 2011, 115, 17925. (357) Valenzano, L.; Civalleri, B.; Sillar, K.; Sauer, J. J. Phys. Chem. C 2011, 115, 21777. (358) Wilmer, C. E.; Snurr, R. Q. Chem. Eng. J. 2011, 171, 775. (359) Babarao, R.; Dai, S.; Jiang, D.-E. Langmuir 2011, 27, 3451. (360) Torrisi, A.; Mellot-Draznieks, C.; Bell, R. G. J. Chem. Phys. 2009, 130, No. 194703. (361) Torrisi, A.; Mellot-Draznieks, C.; Bell, R. G. J. Chem. Phys. 2010, 132, No. 044705. (362) Ramsahye, N. A.; Maurin, G.; Bourrelly, S.; Llewellyn, P. L.; Devic, T.; Serre, C.; Loiseau, T.; Ferey, G. Adsorption 2007, 13, 461. (363) Babarao, R.; Jiang, J. J. Phys. Chem. C 2009, 113, 18287. (364) Salles, F.; Ghoufi, A.; Maurin, G.; Bell, R. G.; Mellot-Draznieks, C.; Ferey, G. Angew. Chem., Int. Ed. 2008, 47, 8487. (365) Dubbeldam, D.; Krishna, R.; Snurr, R. Q. J. Phys. Chem. C 2009, 113, 19317. (366) Jassim, M. S.; Rochelle, G. T. Ind. Eng. Chem. Res. 2006, 45, 2465. (367) Lee, K. B.; Sircar, S. AIChE J. 2008, 54, 2293. (368) An, J.; Geib, S. J.; Rosi, N. L. J. Am. Chem. Soc. 2010, 132, 38. (369) Park, T.-H.; Cychosz, K. A.; Wong-Foy, A. G.; Dailly, A.; Matzger, A. J. Chem. Commun. 2011, 47, 1452. (370) Henke, S.; Fischer, R. A. J. Am. Chem. Soc. 2011, 133, 2064. (371) Cheon, Y. E.; Park, J.; Suh, M. P. Chem. Commun. 2009, 5436. (372) Zelenak, V.; Vargova, Z.; Almasi, M.; Zelenakova, A.; Kuchar, J. Microporous Mesoporous Mater. 2010, 129, 354. (373) Perrin, D. D. Dissociation Constants of Organic Bases in Aqueous Solution: Supplement; Butterworths: London, 1972. (374) Hwang, Y. K.; Hong, D.-Y.; Chang, J.-S.; Jhung, S. H.; Seo, Y.-K.; Kim, J.; Vimont, A.; Daturi, M.; Serre, C.; Ferey, G. Angew. Chem., Int. Ed. 2008, 47, 4144. (375) Yang, Q.; Wiersum, A. D.; Llewellyn, P. L.; Guillerm, V.; Serre, C.; Maurin, G. Chem. Commun. 2011, 47, 9603. (376) Zhao, D.; Yuan, D.; Zhou, H.-C. Energy Environ. Sci. 2008, 1, 222. (377) Sculley, J.; Yuan, D.; Zhou, H.-C. Energy Environ. Sci. 2011, 4, 2721. 779 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW (446) Chae, H. K.; Siberio-Perez, D. Y.; Kim, J.; Go, Y. B.; Eddaoudi, M.; Matzger, A. J.; O’Keeffe, M.; Yaghi, O. M. Nature 2004, 427, 523. (447) Sumida, K.; Hill, M. R.; Horike, S.; Dailly, A.; Long, J. R. J. Am. Chem. Soc. 2009, 131, 15120. (448) Choi, H. J.; Dinca, M.; Long, J. R. J. Am. Chem. Soc. 2008, 130, 7848. (449) Jiang, J. AIChE J. 2009, 55, 2422. (450) Babarao, R.; Eddaoudi, M.; Jiang, J. W. Langmuir 2010, 26, 11196. (451) Babarao, R.; Jiang, J. W. Ind. Eng. Chem. Res. 2011, 50, 62. (452) Van Tassel, P. R.; Davis, H. T.; McCormick, A. V. Langmuir 1994, 10, 1257. (453) Babarao, R.; Jiang, J. W. J. Am. Chem. Soc. 2009, 131, 11417. (454) Yang, Q.; Xu, Q.; Liu, B.; Zhong, C. Chin. J. Chem. Eng. 2009, 17, 781. (455) Kather, A.; Scheffknect, G. Naturwissenschaften 2009, 96, 993. (456) Kakaras, E.; Kounamakos, A.; Doukelis, A.; Giannakopoulos, D.; Vorrias, I. Fuel 2007, 86, 2144. (457) Rezvani, H.; Huang, Y.; McIlveen-Wright, D.; Hewitt, N.; Wang, Y. Fuel 2007, 86, 2134. (458) Greenwood, N. N.; Earnshaw, A. Chemistry of the Elements, 2nd ed.; Butterworth Heinemann: Burlington, MA, 2002. (459) Nandi, S. P.; Walker, P. L., Jr. Sep. Sci. Technol. 1976, 11, 441. (460) Feig, A. L.; Lippard, S. J. Chem. Rev. 1994, 94, 759. (461) Kitajima, A.; Morooka, Y. Chem. Rev. 1994, 94, 737. (462) Magnus, K. A.; Tonthat, H.; Carpenter, J. E. Chem. Rev. 1994, 94, 727. (463) Que, L.; Dong, Y. Acc. Chem. Res. 1996, 29, 190. (464) Sono, M.; Roach, M. P.; Coulter, E. D.; Dawson, J. H. Chem. Rev. 1996, 96, 2841. (465) Karlin, K. D.; Kanderli, S.; Zuberbuhler, A. D. Acc. Chem. Res. 1997, 30, 139. (466) Kovaleva, E. G.; Neibegall, M. B.; Chakrabarty, S.; Lipscomb, J. D. Acc. Chem. Res. 2007, 40, 475. (467) Chufan, E. E.; Puiu, S. C.; Karlin, K. D. Acc. Chem. Res. 2007, 40, 563. (468) Rolff, M.; Schottenheim, J.; Decker, H.; Tuczek, F. Chem. Soc. Rev. 2011, 40, 4077. (469) Ma, S.; Yuan, D.; Chang, J.-S.; Zhou, H.-C. Inorg. Chem. 2009, 48, 5398. (470) Zhong, R.-Q.; Zou, R.-Q.; Du, M.; Yamada, T.; Maruta, G.; Takeda, S.; Li, J.; Xu, Q. CrystEngComm 2010, 12, 677. (471) Chang, Z.; Zhang, D.-S.; Hu, T.-L.; Bu, X.-H. Cryst. Growth Des. 2011, 11, 2050. (472) Collman, J. P.; Del, G.; Guadio, J.; La Mar, G. N.; Balch, A. L. J. Am. Chem. Soc. 1977, 99, 5486. (473) Baldwin, J. E.; Cameron, J. H.; Crossley, M. J.; Dagley, I. J.; Hall, S. R.; Klose, T. J. Chem. Soc., Dalton Trans. 1984, 48, 1739. (474) Sheldon, R. A.; van Santen, R. A. Catalytic Oxidation: Principles and Applications; World Scientific: Singapore, 1995. (475) Dubois, G.; Tripier, T.; Brandes, S. P.; Denat, F.; Guilard, R. J. Mater. Chem. 2002, 12, 2255. (476) Bloch, E. D.; Murray, L. J.; Queen, W. L.; Chavan, S.; Maximoff, S. N.; Bigi, J. P.; Krishna, R.; Peterson, V. K.; Grandjean, F.; Long, G. J.; Smit, B.; Bordiga, S.; Brown, C. M.; Long, J. R. J. Am. Chem. Soc. 2011, 133, 14814. (477) Babcock, H. D.; Herzberg, L. Astrophys. J. 1948, 108, 167. (478) Wong-Foy, A. G.; Matzger, A. J.; Yaghi, O. M. J. Am. Chem. Soc. 2006, 128, 3494. (479) Wu, C.-D.; Hu, A.; Zhang, L.; Lin, W. J. Am. Chem. Soc. 2005, 127, 8940. (480) Seo, J. S.; Whang, D.; Lee, H.; Jun, S. I.; Oh, J.; Jeon, Y. J.; Kim, K. Nature 2000, 404, 982. (481) Chen, B.; Wang, L.; Xiao, Y.; Fronczek, F. R.; Xue, M.; Cui, Y.; Qian, G. Angew. Chem., Int. Ed. 2009, 48, 500. (482) Zhou, X.; Xu, Z.; Zeller, M.; Hunter, A. D. Chem. Commun. 2009, 539. (411) Ma, D.; Li, Y.; Li, Z. Chem. Commun. 2011, 47, 7377. (412) Brandani, F.; Ruthven, D. M. Ind. Eng. Chem. Res. 2004, 43, 8339. (413) Castillo, J. M.; Vlugt, T. J. H.; Calero, S. J. Phys. Chem. C 2008, 112, 15934. (414) Grajciar, L.; Bludsky, O.; Nachtigall, P. J. Phys. Chem. Lett. 2010, 1, 3354. (415) Liu, J.; Benin, A. I.; Furtado, A. M.; Jakubczak, P.; Willis, R. R.; LeVan, M. D. Langmuir 2011, 27, 11451. (416) Higman, C.; van der Burgt, M. Gasification; Elsevier: New York, 2003. (417) Sircar, S.; Golden, T. C. Sep. Sci. Technol. 2000, 35, 667. (418) Damle, A. Hydrogen Fuel: Production, Transport and Storage; CRC Press: Boca Raton, FL, 2004. (419) Hydrogen and Syngas Production and Purification Technologies; Sircar, S., Golden, T. C., Liu, K., Song, C., Subramani, V., Eds.; WileyAIChE: Hoboken, NJ, 2010, p 414. (420) U.S. Department of Energy, National Energy Technology Laboratory: http://www.netl.doe.gov/technologies/carbon_seq/corerd/ pre-combustion.html#sorbents, 2011. (421) Cormos, C.-C.; Starr, F.; Tzimas, E.; Oeteves, S. Int. J. Hydrogen Energy 2008, 33, 1286. (422) Plasynski, S. I.; Litynski, J. T.; McIlvried, H. G.; Srivastava, R. D. Crit. Rev. Plant Sci. 2009, 28, 123. (423) Figueroa, J. D.; Fout, T.; Plasynski, S.; McIlvried, H.; Srivastava, R. D. Int. J. Greenhouse Gas Control 2008, 2, 9. (424) U.S. Department of Energy, National Energy Technology Laboratory: http://www.netl.doe.gov/technologies/carbon_seq/corerd/ pre-combustion.html, 2011. (425) http://www.duke-energy.com/about-us/igcc.asp. (426) Damen, K.; Gnutek, R.; Kaptein, J.; Nannan, N. R.; Oyarzun, B.; Trapp, C.; Colonna, P.; van Dijk, E.; Gross, J.; Bardow, A. Energy Procedia 2011, 4, 1214. (427) Agarwal, A.; Biegler, L. T.; Zitney, S. E. Ind. Eng. Chem. Res. 2010, 49, 5066. (428) Wei, W.; Kulkarni, P.; Liu, K. Hydrogen and Syngas Production and Purification Technologies; Wiley-AIChE: Hoboken, NJ, 2010. (429) U.S. Energy Information Administration: Existing Capacity by Energy Source: www.eia.doe.gov/cneaf/electricity/epa/ epat1p2.html (accessed October 24, 2010). (430) Grande, C. A.; Lopes, F. V. S.; Ribiero, A. M.; Louriero, J. M.; Rodrigues, A. E. Sep. Sci. Technol. 2008, 43, 1338. (431) Lopes, F. V. S.; Grande, C. A.; Ribiero, A. M.; Vilar, V. J. P.; Louriero, J. M.; Rodrigues, A. E. J. Chem. Eng. Data 2010, 55, 184. (432) Kumar, R. Ind. Eng. Chem. Res. 1994, 33, 1600. (433) Ho, M. T.; Alinson, G. W.; Wiley, D. Ind. Eng. Chem. Res. 2008, 47, 4883. (434) Krishna, R.; Long, J. R. J. Phys. Chem. C 2011, 115, 12941. (435) Loulergue, L.; Schilt, A.; Spahni, R.; Masson-Delmotte, V.; Blunier, T.; Lemiuex, B.; Barnola, J.-M.; Raynaud, D.; Stocker, T. F.; Chappellaz, J. Nature 2008, 453, 383. (436) Hamon, L.; Serre, C.; Devic, T.; Loiseau, T.; Millange, F.; Ferey, G.; De Weireld, G. J. Am. Chem. Soc. 2009, 131, 8775. (437) Petit, C.; Mendoza, B.; Bandosz, T. J. ChemPhysChem 2010, 11, 3678. (438) Scholes, C. A.; Smith, K. H.; Kentish, S. E.; Stevens, G. W. Int. J. Greenhouse Gas Control 2010, 4, 739. (439) Zornoza, B.; Martinez-Joaristi, A.; Serra-Crespo, P.; Tellez, C.; Coronas, J.; Gascon, J.; Kapteijn, F. Chem. Commun. 2011, 47, 9522. (440) Chmelik, C.; Heinke, L.; Valiullin, R.; K€arger, J. Chem. Eng. Technol. 2010, 82, 779. (441) Skoulidas, A. I.; Sholl, D. S. J. Phys. Chem. B 2005, 109, 15760. (442) Salles, F.; Jobic, H.; Devic, T.; Llewellyn, P. L.; Serre, C.; Ferey, G.; Maurin, G. ACS Nano 2010, 4, 143. (443) Gallo, M.; Mitnik-Glossman, D. J. Phys. Chem. C 2009, 113, 6634. (444) Liu, J.; Johnson, J. K. J. Low Temp. Phys. 2009, 157, 268. (445) Keskin, S. J. Phys. Chem. C 2010, 114, 13047. 780 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781 Chemical Reviews REVIEW (519) Ordo~ nez, M. J. C.; Balkus, K. J. J.; Ferraris, J. P.; Musselman, I. H. J. Membr. Sci. 2010, 361, 28. (520) Bae, T.-H.; Lee, J. S.; Qiu, W.; Koros, W. J.; Jones, C. W.; Nair, S. Angew. Chem., Int. Ed. 2010, 49, 9863. (521) Basu, S.; Cano-Odena, A.; Vankelecom, I. F. J. J. Membr. Sci. 2010, 362, 478. (522) Hu, J.; Cai, H.; Ren, H.; Wei, Y.; Xu, Z.; Liu, H.; Hu, Y. Ind. Eng. Chem. Res. 2010, 49, 12605. (523) Keskin, S.; Sholl, D. S. Energy Environ. Sci. 2010, 3, 343. (524) Robeson, L. M. J. Membr. Sci. 2008, 320, 390. (483) Shultz, A. M.; Sarjeant, A. A.; Farha, O. K.; Hupp, J. T.; Nguyen, S. J. Am. Chem. Soc. 2011, 133, 13252. (484) Merkel, T. C.; Lin, H.; Wei, X.; Baker, R. J. Membr. Sci. 2010, 359, 126. (485) Gascon, J.; Kapteijn, F. Angew. Chem., Int. Ed. 2010, 49. (486) Zacher, D.; Shekhah, O.; Woll, C.; Fischer, R. A. Chem. Soc. Rev. 2009, 38, 1418. (487) Shekhah, O.; Liu, J.; Fischer, R. A.; Woll, C. Chem. Soc. Rev. 2011, 40, 1081. (488) Scherb, C.; Schodel, A.; Bein, T. Angew. Chem., Int. Ed. 2008, 47, 5777. (489) Horcajada, P.; Serre, C.; Grosso, D.; Boissiere, C.; Perruchas, S.; Sanchez, C.; Ferey, G. Adv. Mater. 2009, 21, 1931. (490) Demessence, A.; Horcajada, P.; Serre, C.; Boissiere, C.; Grosso, D.; Sanchez, C.; Ferey, G. Chem. Commun. 2009, 7149. (491) Yoo, Y.; Jeong, H.-K. Chem. Commun. 2008, 2441. (492) Liu, Y.; Ng, Z.; Khan, E. A.; Jeong, H.-K.; Ching, C.-B.; Lai, Z. Microporous Mesoporous Mater. 2009, 118, 296. (493) Yoo, Y.; Lai, Z.; Jeong, H.-K. Microporous Mesoporous Mater. 2009, 123, 100. (494) Guo, H.; Zhu, G.; Hewitt, I. J.; Qiu, S. J. Am. Chem. Soc. 2009, 131, 1646. (495) Ranjan, R.; Tsapatsis, M. Chem. Mater. 2009, 21, 4920. (496) Bux, H.; Liang, F.; Li, Y.; Cravillon, J.; Wiebcke, M.; Caro, J. J. Am. Chem. Soc. 2009, 131, 16000. (497) Bux, H.; Chmelik, C.; van Baten, J. M.; Krishna, R.; Caro, J. Adv. Mater. 2010, 22, 4741. (498) Bux, H.; Feldhoff, A.; Cravillon, J.; Wiebcke, M.; Li, Y.-S.; Caro, J. Chem. Mater. 2011, 23, 2262. (499) Li, Y.-S.; Bux, H.; Feldhoff, A.; Li, G.-L.; Yang, W.-S.; Caro, J. Adv. Mater. 2010, 22, 3322. (500) Li, Y.-S.; Liang, F.-Y.; Bux, H.; Feldhoff, A.; Yang, W.-S.; Caro, J. Angew. Chem., Int. Ed. 2010, 49, 548. (501) Li, Y.; Liang, F.; Bux, H.; Yang, W.; Caro, J. J. Membr. Sci. 2010, 354, 48. (502) Huang, A.; Bux, H.; Steinbach, F.; Caro, J. Angew. Chem., Int. Ed. 2010, 49, 4958. (503) Huang, A.; Dou, W.; Caro, J. J. Am. Chem. Soc. 2010, 132, 15562. (504) Venna, S. R.; Carreon, M. A. J. Am. Chem. Soc. 2010, 132, 76. (505) McCarthy, M. C.; Varela-Guerrero, V.; Barnett, G. V.; Jeong, H.-K. Langmuir 2010, 26, 14636. (506) Yao, J.; Dong, D.; Li, D.; He, L.; Xu, G.; Wang, H. Chem. Commun. 2011, 47, 2559. (507) Snyder, M. A.; Tsapatsis, M. Angew. Chem., Int. Ed. 2007, 46, 7560. (508) Choi, J.; Jeong, H.-K.; Snyder, M. A.; Stoeger, J. A.; Masel, R. I.; Tsapatsis, M. Science 2009, 325, 590. (509) Lai, Z.; Bonilla, G.; Diaz, I.; Nery, J. G.; Sujaoti, K.; Amat, M. A.; Kokkoli, E.; Terasaki, O.; Thompson, R. W.; Tsapatsis, M.; Vlachos, D. G. Science 2003, 300, 456. (510) Takamizawa, S.; Takasaki, Y.; Miyake, R. J. Am. Chem. Soc. 2010, 132, 2862. (511) Krishna, R.; van Baten, J. M. J. Membr. Sci. 2011, 377, 249. (512) Aguado, S.; Nicolas, C.-H.; Moizan-Basle, V.; Nieto, C.; Amrouche, H.; Bats, N.; Audebrand, N.; Farrusseng, D. New J. Chem. 2011, 35, 41. (513) Won, J.; Seo, J. S.; Kim, J. H.; Kim, H. S.; Kang, Y. S.; Kim, S. J.; Kim, Y.; Jegal, J. Adv. Mater. 2005, 17, 80. (514) Car, A.; Stropnik, C.; Peinemann, K.-V. Desalination 2006, 200, 424. (515) Zhang, Y.; Musselman, I. H.; Ferraris, J. P.; Balkus, K. J. J. J. Membr. Sci. 2008, 313, 170. (516) Perez, E. V.; Balkus, K. J. J.; Ferraris, J. P.; Musselman, I. H. J. Membr. Sci. 2009, 328, 165. (517) Adams, R.; Carson, C.; Ward, J.; Tannenbaum, R.; Koros, W. Microporous Mesoporous Mater. 2010, 131, 13. (518) Diaz, K.; Garrido, L.; Lopez-Gonzalez, M.; del Castillo, L. F.; Riande, E. Macromolecules 2009, 43, 316. 781 dx.doi.org/10.1021/cr2003272 |Chem. Rev. 2012, 112, 724–781