Sensing and capture of toxic and hazardous gases and vapors by metal–organic frameworks

Hao Wang , William P. Lustig and Jing Li *
Department of Chemistry and Chemical Biology, Rutgers University, Piscataway, NJ 08854, USA. E-mail:

Received 19th December 2017

First published on 13th March 2018

Toxic and hazardous chemical species are ubiquitous, predominantly emitted by anthropogenic activities, and pose serious risks to human health and the environment. Thus, the sensing and subsequent capture of these chemicals, especially in the gas or vapor phase, are of extreme importance. To this end, metal–organic frameworks have attracted significant interest, as their high porosity and wide tunability make them ideal for both applications. These tailorable framework materials are particularly promising for the specific sensing and capture of targeted chemicals, as they can be designed to fit a diverse range of required conditions. This review will discuss the advantages of metal–organic frameworks in the sensing and capture of harmful gases and vapors, as well as principles and strategies guiding the design of these materials. Recent progress in the luminescent detection of aromatic and aliphatic volatile organic compounds, toxic gases, and chemical warfare agents will be summarized, and the adsorptive removal of fluorocarbons/chlorofluorocarbons, volatile radioactive species, toxic industrial gases and chemical warfare agents will be discussed.

image file: c7cs00885f-p1.tif

Hao Wang

Hao Wang obtained his BS from Wuhan University in China in 2012. He then joined the Jing Li research group at Rutgers University and is now in his fifth year of the PhD program. His research focuses on the design and synthesis of microporous metal–organic frameworks and their applications in gas adsorption and separation, with an emphasis on hydrocarbon separation, noble gas separation, and capture of hazardous gases and vapors.

image file: c7cs00885f-p2.tif

William P. Lustig

William P. Lustig obtained his BS in chemistry from American University in 2010, joined the Jing Li Research Group in 2013, and is currently pursuing a PhD in chemistry under Professor Jing Li's guidance. His research is centered on the synthesis and development of new organic chromophore-based luminescent metal organic frameworks with applications as sensor and phosphor materials. He is especially interested in studying the luminescence mechanisms at play in these systems and using theoretical methods to aid in their rational design.

image file: c7cs00885f-p3.tif

Jing Li

Jing Li is a Distinguished Professor in the Department of Chemistry and Chemical Biology at Rutgers University. She received her PhD degree from Cornell University in 1990 under the guidance of Professor Roald Hoffmann. She joined the chemistry faculty at Rutgers University in 1991 as Assistant Professor. She was promoted to Associate Professor in 1996, to Full Professor in 1999, and to Distinguished Professor in 2006. She has published +320 research articles, book chapters, and invited reviews, and holds 13 issued and pending patents. She was recognized as “Highly Cited Researcher” by Thomson Reuters in both 2015 and 2016.

1. Introduction

Metal–organic frameworks or MOFs, are coordination networks containing potential voids, following IUPAC recommendations.1 Owing to their fascinating structural chemistry and enormous potential in industrial applications, MOFs have drawn tremendous attention over the past two decades from both scientific researchers and industrial engineers. As MOFs are built by coordinative bonds between metal nodes (metal ions or clusters) and organic linkers, a nearly infinite number of MOFs can be achieved by altering their connectivity or changing the identity of either metal or ligand. The unique features of MOFs include exceptionally high porosity (BET surface area up to 7000 m2 g−1, and pore volume up to 4.4 cm3 g−1),2 compositional and structural diversity, and highly tunable pore shape/size and surface functionality, to name a few.3 Fundamental studies with respect to the coordination, connectivity, and topology of MOFs have greatly enriched the knowledge and extended the horizon of chemists in the field, while the exploration of MOFs for industrial applications has continued to accelerate.4,5 As a family of multifunctional materials, MOFs have been extensively studied for various potential applications including gas storage,6–9 molecular separations,10–14 catalysis,15–18 chemical sensing,19–22 proton conductivity,23–26 and many others.27,28 In some of these areas MOFs have outperformed traditional or benchmark materials, or have shown potential value for commercialization. For example, a recently reported microporous MOF is capable of separating propane and propylene through selective molecular exclusion which is not achievable by traditional zeolite materials.29 More recently, the chemical company BASF developed a MOF-based natural gas storage system and have been testing it in demonstration vehicles, indicating that this technology is getting close to the market.30 Additionally, TruPick, a post-harvest freshness management tool for fruits and vegetables built on MOF adsorbent, has already been used commercially in the United States. The technology uses MOFs for the storage and release of 1-methylcyclopropene (1-mcp), with the goal of prolonging the time over which fruits and vegetables can be safely stored.30 There are a great many other examples wherein MOFs have shown enormous promise for implementation in real-world systems, particularly those associated with issues concerning energy and the environment. Among these is the sensing and capture of hazardous gases and vapors.

1.1. Hazardous gases and vapors, sources and importance of sensing and capture

Hazardous gases and vapors, including but not limited to toxic industrial gases (COx, NH3, SOx, NOx, H2S etc.), volatile organic compounds (VOCs, such as hydrocarbons, fluorocarbons, chlorofluorocarbons, etc.), volatile radioactive species, and chemical warfare agents, are a major threat to human health and the environment.31–33 These hazardous gases and vapors are mainly released into the atmosphere from anthropogenic sources including power plants, factories, and household emissions, to name a few. For example, the sharply rising level of atmospheric carbon dioxide is predominantly attributed to the combustion of coal, oil, and natural gas which accounts for 80% of the CO2 emission worldwide.34,35 In addition, the emission of volatile radioactive species such as iodine and organic iodides is primarily associated with the implementation of nuclear power.36,37 Chemical warfare agents such as sarin and sulfur mustard have been frequently used in localized conflicts and terrorist attack.38 Toxic industrial gases are ubiquitous in industrial processes, and chemical workers or related personnel are at risk of exposure in case of any accidental spillage or leakage; chlorofluorocarbons, mostly emitted from the use of refrigerants, are responsible for the depletion of ozone layer.39 In light of the impacts that hazardous gases and vapors have on human health and the environment, developing effective technologies for the sensing and capture of toxic chemicals and environmental pollutants are therefore of global importance and highly necessary. Advanced sensor materials will enable fast detection of the presence of toxic or hazardous species, and adsorbent materials that can effectively capture toxic and hazardous gases and vapors are vital for their removal and subsequent sequestration.

1.2. Luminescent MOFs (LMOFs) as chemosensors, mechanisms of detection, advantages and general strategies

In luminescent sensing of gases and vapors, the presence of a given analyte is detected through the modulation of luminescence from a probe material. This typically involves emission turn-on, emission turn-off, or shifts in the emission energy/wavelength from the luminescent probe. This type of sensing is advantageous in that it combines technical simplicity with the potential for extremely powerful performance. The instrumentation required only consists of an excitation source, probe material, emission detector, and signal output. The resulting devices can be extremely cost effective, and depending on the specific application, can often be compact enough for mobile use. Moreover, despite their low cost, small size, and ease of use, selective ppb-level sensitivity can be achieved through careful design of the probe material.

An effective luminescent probe should have short response time, good sensitivity and selectivity for the analyte, strong emission when in the on-state, high stability and reusability in real-world conditions. Because of their exceptional tunability, luminescent metal–organic frameworks (LMOFs) are especially effective in this role. Through the alteration of metal ions, organic ligands, guest molecules, and conditions used in synthesis, plus post-synthetic modification, nearly every physical and chemical quality of an LMOF can be tuned.

Porosity, pore geometries, and pore surface chemistry can be controlled to maximize selective interactions between the framework and the analyte material. This allows for the sensitivity, selectivity, and recyclability of the probe to be optimized. Simple adjustments of the pore dimensions allow for size-based selectivity. This can be accomplished by adjusting the actual pore size and geometry, or by partially occluding the pore through functionalization of the inner surface or the inclusion of guest molecules.40 Similarly, controlling the chemical environment of the pore through ligand design, ligand functionalization, or specific guest inclusion can allow for the selection of species by their chemical properties. Through the use of hydrophobic ligands, for example, hydrophilic molecules may be excluded from the pores of the material, further increasing its selectivity for a given hydrophobic analyte.41 In addition to tuning the broad chemical environment, specific functional groups that interact strongly with the desired analyte may be included in the pore to enhance selectivity for that material. Ligands with Lewis-basic moieties, such as amine-based functional groups, can be used to increase interactions with Lewis-acidic analytes.42 Post-synthetic removal of terminal ligands may expose open metal sites, allowing for the coordination of Lewis-basic analytes.43 Ligands with large, planar, aromatic regions can increase π–π stacking interactions between the framework and aromatic analyte molecules.44 These optimizations not only impact selectivity, but sensitivity as well. By improving the ability of an LMOF to selectively interact with the analyte material, preconcentration of the analyte within the LMOF can be achieved.45 This increases the local concentration of the analyte, allowing for extremely efficient sensing even when the general concentration of the analyte might otherwise be too low to detect.

LMOFs are also well suited as luminescent probe materials because of their favourable luminescence qualities. For example, in the cases of rigid frameworks built on organic chromophores, non-radiative excitation decay pathways can be significantly reduced compared with the chromophores in form of free molecules, resulting in LMOFs with extremely strong emission and quantum yields approaching unity.46,47 Additionally, their multi-component design introduces a variety of potential emission mechanisms. In ligand-centered (LC) emission, excitation and emission processes are both located on a single ligand molecule, whereas ligand-to-ligand charge transfer (LLCT) involves the transfer of electron density from a donor ligand to an acceptor ligand upon excitation, with the reversal up emission. Similarly, metal-to-ligand charge transfer (MLCT) or ligand-to-metal charge transfer (LMCT) can occur, with the same movement of electron density from donor to acceptor upon excitation. Fluorescence resonance energy transfer (FRET) can also play a role in MOF luminescence. Excitation energy is first absorbed by a donor species—typically a ligand molecule—after which dipole–dipole interactions with the acceptor species—either another ligand or a metal—permits the non-radiative transfer of excitation energy from the donor to the acceptor. For this to occur, the donor must first possess an emissive transition with energy exactly matching that of an excitation transition in the acceptor, giving rise to spectral overlap between the emission of donor species and absorbance of acceptor species. Second, the dipole moments of the donor emission transition and acceptor excitation transition must be non-orthogonal. Following this energy transfer, the excited acceptor species emits.

A subclass of this energy transfer mechanism, or sensitization, forms the basis of emission in the expansive field of lanthanide-based LMOFs. Most trivalent lanthanide ions possess sharp, characteristic emission profiles, and they can be included either in the SBU or post-synthetically added to the material's pores. However, as direct excitation of lanthanide ions requires parity-forbidden f–f transitions, their absorbance is extremely weak. To overcome this challenge, sensitizing ligands are used. Upon exposure to excitation energy, an electron is excited into the singlet excited state S1,S on the ligand, after which it undergoes intersystem crossing into the triplet excited state T1,S on the ligand. Energy transfer moves the excited state from the ligand T1,S to the emissive lanthanide triplet state T1,A, where a photon is emitted.48

Until now, these charge transfer and energy transfer mechanisms have been discussed as functioning between ligand molecules and metal ions. However, the inherent porosity of MOFs permits guest molecules within the pores to participate in these mechanisms as well. Guest molecules can act as independent luminescence centers, or they can participate in charge transfer to and from ligands or metals. Finally, multiple emission mechanisms can occur simultaneously within a single LMOF.

In designing an LMOF-based probe material to take advantage of these mechanisms, many options are available. As an analyte interacts with an LMOF it may induce changes in the emission wavelength, emission quenching, or emission enhancement. Shifting the emission energy occurs when interactions between the analyte and LMOF alter the LMOF electronic structure.49 This can be accomplished through including functionality that directly interacts with the targeted analyte as described above. Additionally, it is possible to take advantage of vapochromic behavior in an LMOF, where the adsorption of polar or nonpolar molecules can stabilize or destabilize the excited state and thereby alter LMOF excited state energy levels and the wavelength of resulting emission.50

The most common quenching mechanisms are charge transfer, in which a photoexcited electron is transferred from the higher-lying LUMO of the LMOF into the lower-lying LUMO of the (typically electron-deficient) analyte, and FRET, in which overlap between the emission spectra of the LMOF and absorbance spectra of the analyte permits transfer of excitation energy from the LMOF to the analyte, where it decays non-radiatively. While direct orbital overlap between LMOF and analyte molecules is required for charge transfer, FRET can take place over longer distances on the nanometer scale, and so requires only that the analyte be present in or near the LMOF's pore.48 Emission enhancement can occur through a similar mechanism, with photoexcited electrons from the higher-lying analyte LUMO transferred into the lower-lying LMOF LUMO.

In either case, modulation of emission intensity from the LMOF requires specific relationships between the LUMO energy levels of the target analyte and LMOF, or spectral overlap between the LMOF emission and analyte absorbance. Using a chromophoric ligand-based strategy to prepare LMOFs with LC-based emission is useful method of designing an LMOF sensor with the appropriate LUMO energy levels or emission wavelength to interact with a targeted analyte.51 This strategy entails preparing an emissive ligand based on an organic chromophore with the optoelectronic properties and functionality necessitated by the target analyte, then constructing it into a MOF with d0 or d10 transition metals. These closed d subshell species have relatively low-lying HOMOs (highest occupied molecular orbitals) and high-lying LUMOs that usually preclude their participation in luminescence. The resulting MOF should then possess similar properties (emission and excitation spectra, HOMO and LUMO energy levels) as the initial chromophore.

Lanthanide LMOFs that participate in sensitized emission provide another lever by which an analyte can affect emission from the material. By interacting with the sensitizing ligand, an analyte can increase or decrease the efficiency of sensitization, leading to enhancement or quenching of emission from the emissive lanthanide.52 And while these strategies for preparing sensor materials are the most common, this is not an exhaustive list. Because of their great flexibility and tunability, creative researchers can induce changes in luminescence using any number of methods. Rationally designed materials that exhibit luminescence changes as a result of gas-adsorption-induced breathing have been reported,53 as well as materials with emission turn-on when the adsorbed analyte displaced emission-quenching atmospheric O2.54 Others have reported selective hydrogen-bonding analytes that rigidify unbound pendant functional groups and thereby enhance emission,55 or the selective oxidation of an analyte species that results in enhanced emission from the LMOF.45 Regardless, the exceptional properties of MOFs as luminescent probe materials promise that the rapid growth in the field will continue.

1.3. Design considerations of MOFs for the capture of hazardous gases and vapors

Adsorption related applications that make use of the porosity of MOFs are the most extensively studied areas among various aspects of MOF materials. MOFs are particularly promising for the capture and removal of target species because of their high porosity and tunability, which may not be readily achievable for traditional adsorbent materials.56,57 This removal of the target species can be accomplished via bulk sequestration within the MOF structure followed by later desorption, or occasionally by the catalytic decomposition of the species within the MOF pore. The adsorption and decomposition of a target molecule has been extensively reported in the solution phase,58–60 and while it isn’t as common in the vapor/gas phase, some examples do exist.61 However, as this review will primarily discuss and adsorption of toxic and hazardous gases, the decomposition of adsorbed species will not be a focus of this review.

To some extent, tailor-made MOFs with desirable pore shape, pore size, and surface functionality are attainable by design. For example, by applying reticular chemistry and ligand functionalization, one can fine-tune pore size and surface properties of MOFs for specific applications.62–64

The capture of hazardous gases and vapors commonly involves selective adsorption of target molecules from a mixture, usually under relatively low concentration. Within this context, various parameters including adsorption selectivity, uptake capacity, stability, recyclability, and cost should be taken into consideration while evaluating the performance of an adsorbent candidate.

Firstly, it is important to note that adsorption selectivity and uptake capacity often have an inverse relationship, as selective adsorption is favored by relatively smaller pores for maximum size discrimination and/or sufficient adsorbate–adsorbent contact, while a high uptake capacity is typically favored by highly porous adsorbents.65,66 However this is not always the case, as both qualities also depend on the pore structure and surface functionality of the adsorbents, as well as the physical/chemical features of the adsorbates. For example, we recently reported the use of MIL-101-Cr based molecular traps for the capture of radioactive organic iodides from nuclear waste.67 The tertiary amine functionalized MIL-101-Cr materials are able to selectively adsorb radioactive organic iodides with both high uptake capacity and selectivity, attributed to the high porosity of MIL-101-Cr and the tailor-made surface functionality which enables a specific interaction between the target adsorbate and the adsorbent. This example demonstrates how MOFs possessing high porosity and desirable functionality are advantageous for the capture of hazardous gases and vapors at low concentration. Thus, functionalization of existing, prototype materials represents an effective way to improve the capture performance.68 For example, the amino-functionalized MOF-5 and hydroxyl-functionalized MOF-205 show substantially enhanced ammonia capture capability compared to their parent compounds, owing to the formation of strong hydrogen bonds between ammonia and functional groups from the adsorbents.69,70 The same strategy has proven effective for the capture of SOx, NOx, H2S etc.71

Stability (both thermal and chemical stability) is another crucial factor that influences the performance of an adsorbent material. It has commonly been neglected in the early exploration of MOFs, especially with regards to water/chemical stability. It has been shown that the gas capture capability of MOFs may drop significantly under real-world systems or simply in the presence of moisture, despite previous adsorption measurements performed under dry conditions that indicate very high performance.72 But over the past few years, the stability of MOFs has been greatly enhanced through the use of high valence metals such as zirconium, aluminum, and yttrium.73–75 Some of these MOFs have exhibited exceptional framework robustness which can be thermally stable up to 400+ °C and retain their crystallinity and porosity in hot water or even in acidic or basic solutions.76,77 Additionally, the introduction of hydrophobic linkers can improve the water stability of MOFs, and with the added benefit of depressing the competitive adsorption of moisture.78–80

Recyclability, which is normally correlated with cost, must also be taken into account when considering an adsorbent material for industrial implementation. MOFs are typically reusable in cases where the capture process involves only physisorption without altering the integrity of the adsorbent. When chemisorption is involved, the adsorbent is usually unrecyclable. However, it has been shown that chemisorbed species may be fully desorbed under optimized conditions, making the adsorbent recyclable.81 Additionally, for post-synthetically functionalized MOFs, it is possible to remove the chemisorbed adsorbates together with the functional moieties and reuse the original adsorbents.67

Finally, the cost of a MOF is usually dominated by the organic linker used, as the most common MOF node metals (Zn, Cu, Zr, Al etc.) are earth abundant and inexpensive. Thus, for real-world applications, low cost and readily available ligands, rather than those synthesized through complicated organic reactions, are favorable.

2. Detection of harmful gases and vapors by LMOFs

Pollution of the air, soil, and water is a global issue, with diseases resulting from pollution responsible for 9 million premature deaths in 2015, or 16% of all deaths worldwide.82 The treatment of pollution-related costs is also a burden on health systems, with welfare losses due to pollution accounting for 6.2% of global economic output.82 Pollution of the atmosphere by gas and vapor-phase chemical species is contributes the majority of this risk, with approximately 7 million deaths attributable to air pollution in 2012.83 Industrial and power plant exhaust streams, vehicle exhaust, outgassing from materials, and improper waste disposal all play key roles in introducing these harmful species into the atmosphere, and monitoring their concentration is a key component of any assessment of air quality. Additionally, the detection of specific gasses and volatile compounds is of use in industrial safety monitoring. LMOFs provide an excellent opportunity to develop new, cost-effective alternatives to existing detection methods.

2.1. Detection of aromatic VOCs

Aromatic VOCs include species like benzene, toluene, nitroaromatic species, aromatic amines like aniline, and many other compounds. They are commonly used in myriad ways by various industry, from solvents and coatings to pesticides, medicinal precursors, packaging, and building materials. Vehicle exhaust is another common source of aromatic VOCs. Some, such as benzophenone or benzaldehyde, have limited toxicity and are commonly used as flavoring agents or in soaps. Others, like benzene, are acutely toxic and carcinogenic.86 Exposure to hazardous aromatic VOCs can be an occupational hazard for employees working in industries that employ them and an environmental hazard for those living or working near waste-disposal sites. Exposure can also arise via outgassing of the aromatic VOC from building materials such as particleboard or flooring adhesives.87

Nitroaromatic compounds (NACs) were among the first class of VOCs to be detected in the vapor phase by LMOF materials. In 2009, we reported Zn2(bpdc)2(bpee) (bpdc = 4,4′-biphenyldicarboxylate, bpee = 1,2-bipyridylethene), a porous and strongly LMOF with blue emission.88,89 In the presence of dinitrotoluene (DNT) or nitrobenzene (NB) vapors at approximately 0.18 ppm and 300 ppm, respectively, emission from the LMOF was strongly and rapidly quenched, with an 85% reduction in luminescence intensity following a 10 second exposure (Fig. 1). This was accomplished through a redox quenching mechanism. Following photoexcitation, excited LMOF electrons were transferred from the lowest unoccupied molecular orbital (LUMO) of the LMOF into the lower-lying LUMO of the electron-deficient nitroaromatic compound, quenching emission from the LMOF. This sensing behavior was fully reversible by heating the sensor LMOF at 150 °C for about one minute (Fig. 1).

image file: c7cs00885f-f1.tif
Fig. 1 Graph showing the emission quench percentage vs. time for the LMOF Zn2(bpdc)2(bpee) following exposure to DNT vapors. The structure of DNT is inset, along with the LMOF's emission spectra before exposure (emission peak at 420 nm) and after exposure (emission peak at 462 nm). Recyclability tests are also inset, showing the intensity of emission before (dark grey) and after (light grey) exposure to DNT vapors over several cycles.88 Reproduced from ref. 88 with permission from the John Wiley & Sons, Inc., copyright 2009.

A follow-up work of ours introduced the ability to identify the analyte through the use of 2D signal modulation, with exposure to different analyte NACs altering both emission intensity and emission wavelength to different degrees.90 The LMOF Zn2(ndc)2(bpe) (ndc = 2,6-naphthalenedicarboxylate, bpe = 1,2-bis(4-pyridyl)ethane) was exposed to vapors of NB, 2-nitrotoluene (mNT), 1,3-dinitrobenzene (mDNB), 2,4-dinitrotoluene (2,4-DNT), 1,4-dinitrobenzene(pDNB), and 2,4,6-trinitrotoluene (TNT), in addition to a number of other solvent molecules. Following a 5 minute exposure, emission from the LMOF was quenched by 45–95% and blueshifted by 2–25 nm, depending on the analyte in use. Molecular loading simulations and DFT calculations were used to demonstrate that the strength of interaction between the analyte and framework was responsible for the degree of emission blue shift.

As discussed earlier, improving the stability of LMOF sensor materials in the presence of moisture was a vital requirement for real-world application. An example of a moisture-stable LMOF able to detect NAC vapors was reported in 2015 by Zang and Hou et al.91 using the metal ion Tb3+. When exposed to NB and mNT vapors, the characteristic Tb3+ emission from the LMOF Tb(L)(OH) (L = 5-(4-carboxyphenyl)pyridine-2-carboxylate) was quenched. Emission from the LMOF in the absence of analyte followed the standard mechanism for sensitized lanthanide systems—excitation of the ligand singlet state, followed by intersystem crossing into the ligand excited triplet state, and finally energy transfer to the lanthanide triplet state that resulted in characteristic lanthanide emission. However, in the presence of NACs, the initial excited electron was instead transferred from the ligand singlet state into the lower-lying LUMO of the electron deficient analyte, quenching emission from the LMOF.

Aromatic amines are another aromatic VOC pollutant with significant health hazards. While most vapor-phase NAC detection involves π–π interactions between the analyte and framework, the amine group present in aniline and other aromatic amines provides an additional target for interaction. An example of an LMOF with hydrogen-bond acceptor moieties for aniline detection was reported in 2014 by Zhao and Li et al.92 The cadmium-based LMOF [CdL]·[H2N(CH3)2]+(DMF)(H2O)3, synthesized using the amide-containing ligand bis(3,5-dicarboxyphenyl)terephthalamide (H4L), emits strong blue light at 450 nm through a ligand-centered emission process. Following exposure to aniline vapors, emission from the LMOF was quenched by 15% after 200 seconds, with the majority of the quenching occurring after only 25 seconds. The authors ascribe interaction between the LMOF and the aniline vapor to both π–π interactions between the aromatic moieties on the ligand and analyte and hydrogen bonding between the aniline amine hydrogen and the amide group in the ligand molecule.

Amine basicity also provides a method of amine-specific sensing, as demonstrated by Lin and Huang et al.44 The electron-poor ligand DPNDI (N,N′-di(4-pyridyl)-1,4,5,8-naphthalene diimide) was first reacted with AnSiF6 clusters to produce a weakly emissive LMOF, which was then loaded with electron-rich naphthalene guests to create 1a⊃naphthalene. This guest-loaded LMOF fluoresced brightly at 600 nm due to an exciplex electronic charge transfer state between the electron-rich and the electron-poor framework ligand. When 1a⊃naphthalene was exposed to basic vapors, in addition to a number of aliphatic amines, electron transfer from the strongly basic vapor to the π-acidic ligand interrupted the emissive charge transfer mechanism and quenched emission. The interruption of the exciton formation also resulted in a clear color change (Fig. 2). Finally, the guest-loaded LMOF also demonstrated some size preference, with the bulkiness of the amine inversely related to the strength of the quenching interaction.

image file: c7cs00885f-f2.tif
Fig. 2 Color change induced in 1a⊃naphthalene when exposed to amine vapors.44 Reproduced from ref. 44 with permission from the American Chemical Society, copyright 2016.

While the majority of luminescent sensors exhibit emission turn-off in the presence of a given analyte, emission turn-on is also possible, and is often preferred as it is less susceptible to false positive signals related to device malfunction. As quenching is often observed in cases of electron-deficient analytes with LUMO energy levels below that of the sensor LMOF, emission enhancement can occur in electron-rich analytes with LUMO energy levels above that of the LMOF sensor. We reported an early example of this interaction in 2011, with the LMOF Zn2(oba)2(bpy) (oba = 4,4′-oxybis(benzoic acid), bpy = 4,4′-bipyridine).93 While it demonstrated emission quenching in the presence of NAC vapors, the LMOF emission at 420 nm was enhanced in the presence of electron-rich aromatic VOCs benzene (80% enhancement), chlorobenzene (70% enhancement) and toluene (120% enhancement) (Fig. 3). The degree of increase in emission intensity was in trend with the electron density in the benzene ring, and DFT calculations indicated that the LUMO of these three analytes was indeed higher than that of the LMOF. Cyclic voltammetry measurements showed that the reduction potential of the three analytes were more negative than the LMOF, indicating that the LMOF would act as an electron acceptor. Our subsequent studies have shown such electron transfer is a very common process observed in MOFs.94–96

image file: c7cs00885f-f3.tif
Fig. 3 (left) Enhancement of emission from the LMOF Zn2(oba)2(bpy) following exposure to toluene vs. time of exposure, with before/after exposure emission spectra and recyclability chart inset. (right) Enhancement of LMOF emission in the presence of toluene (TO), benzene (BZ), and chlorobenzene (ClBZ) vapor.93 Reproduced from ref. 93 with permission from the American Chemical Society, copyright 2011.

A similar sensor for benzene with a much stronger response was recently reported by Lan and Sun et al.97 The porous cadmium-based LMOF has the formula Cd3(L)(bipy)2·4H2O (H6L = (tri-((4-carboxyphenoxy)methyl)methoxy)-tri-((4-carboxyphenoxy)methyl)methane) and exhibits approximately an 8-fold increase in luminescence intensity when exposed to benzene vapors with a response time of less than one minute. The LMOF fluoresces at 381 nm under 314 nm excitation through a ligand-to-ligand charge transfer process. DFT calculations indicated that the lowest unoccupied molecular orbital (LUMO) of the LMOF was primarily located on the aromatic bipy ligand. These calculations also indicated that the LUMO of benzene is slightly higher than the LUMO of the LMOF, allowing energy transfer from benzene to the LMOF LUMO to enhance the emission intensity. Exposure to nitrobenzene, conversely, caused a strong quenching response, as calculations showed the LUMO of nitrobenzene to be lower than that of the LMOF, resulting in energy transfer out of the excited LMOF LUMO and into that of nitrobenzene, as in previous examples (Fig. 4).

image file: c7cs00885f-f4.tif
Fig. 4 (left) Representative fragment of the LMOF Cd3(L)(bipy)2·4H2O used in DFT calculations, with molecular orbitals corresponding to the fragment HOMO and LUMO shown. (right) Schematic demonstrating the relative positions of the calculated LMOF fragment, benzene, and nitrobenzene HOMOs and LUMOs, as well as the proposed mechanism of emission quenching or enhancement.97 Reproduced from ref. 97 with permission from the Royal Society of Chemistry, copyright 2015.

2.2. Detection of aliphatic VOCs

Aliphatic VOCs are another class of common atmospheric pollutant, with anthropogenic sources accounting for the emission of approximately 142 million metric tons of VOC carbon per year.98 Much like aromatic VOCs, these compounds are widely used as solvents or additives in paints, coatings, polymers, building materials, office equipment, and fuels. They typically enter the atmosphere through evaporation, outgassing, or following the incomplete combustion of fossil fuels. Atmospheric VOC levels have been shown to be elevated 2–5 fold in indoor residential spaces when compared to outdoor spaces, regardless of the rural or urban location of the space, with some aliphatic VOCs posing serious health risks. In industrial settings that utilize these VOCs, the levels of exposure can be even higher.

Yan and Xu reported an interesting LMOF composite material for the sensitive and selective determination of aliphatic aldehyde vapors, with the specific application of detecting aldehyde pollution in automobiles.45 10 nm ZnO nanoparticles were prepared, then reacted with ZrCl4 and H2bpydc (2,2′-bipyridine-4,4′-dicarboxylic acid) to form a UiO-66-type MOF (UiO-MOF) around the ZnO nanoparticles, which was confirmed by TEM and spectroscopic studies. The resulting ZnO@UiO-MOF (ZUM) composite material was then loaded with Eu3+ post-synthetically to generate Eu@ZUM, with the Eu3+ atoms coordinated to the bipyridine moiety of the bpydc ligands. Under 365 nm excitation, emission from Eu@ZUM was a mix of ligand-centered emission at 470 nm and Eu-centered emission at 590, 614, and 700 nm, with the intensity at 614 nm: the intensity at 470 nm (I614/I470) = 2.3. The Eu@ZUM was mounted onto a strip of test paper and exposed to a series of vehicle cabin pollutants, including formaldehyde (FA) benzene (Ben); ortho-, meta-, and para-xylene (OX, MX, and PX); ethylbenzene (EB); butyl acetate (BA); toluene (Tol); and cyclohexane (CH). The resulting I614/I470 was 2.3 ± 0.1 for all analytes except FA, which had I614/I470 = 5.5 (Fig. 5). Subsequent trials with larger aldehydes acetaldehyde (AA) and acraldehyde (ACA) also gave elevated I614/I470 of 4.3 and 3.3, respectively. Additionally, the linear relationship between the concentration of FA vapor and I614/I470 permitted the authors to calculate a limit of detection (LOD) of 42 ppb for FA at 25 C. Finally, the authors demonstrated that the sensing was completely reversible by removing the test paper from the FA-containing atmosphere. When investigating the sensing mechanism, it was found that the lifetime of the Eu-centered emission at 614 was unchanged in the presence of FA; the authors therefore ruled out direct interaction between FA and Eu3+. Additionally, since the intensity of the ligand-centered emission at 470 nm was consistent with the other analytes (Fig. 5), the authors ruled out some interaction between FA and bpydc that increased the efficiency of sensitization. Instead, it was determined that increased electron density in the valence band of the ZnO nanoparticles within Eu@ZUM—caused by energy transfer from the excited UiO-MOF to the ZnO nanoparticles—ionizes preadsorbed molecular oxygen, which oxidizes the aldehyde analyte molecule, with the emancipated electrons being injected into the Eu3+ excited state and resulting in enhance emission.

image file: c7cs00885f-f5.tif
Fig. 5 (top) Emission spectra of Eu@ZUM under 365 nm excitation when exposed to VOC pollutant vapors. (bottom) Relative intensity of emission from Eu@ZUM at 614 nm compared to emission at 470 nm when exposed to pollutant vapors, with an image of the test strips illuminated by 365 nm UV light inset. Abbreviations: FA = formaldehyde, Ben = benzene, PX = para-xylene, MX = meta-xylene, OX = ortho-xylene, EB = ethylbenzene, BA = butyl acetate, Tol = toluene, CH = cyclohexane, origin = native Eu@ZUM.45 Reproduced from ref. 45 with permission from the Royal Society of Chemistry, copyright 2017.

A pair of LMOFs able to detect both amine and aldehyde vapors was reported in 2016 by Yang, Song, and Ma et al.100 Both LMOFs used an unusual chair-conformation resorcin[4]arene-based octacarboxylate ligand H8L, with 1 having the formula [Cd2(L)][(CH3)2NH2]4·4H2O and 2 having the formula [Zn2(L)][(CH3)2NH2]4+·2DMF·4H2O. Both exhibited blue-green ligand-based emission, which was quenched following short exposure to amine and aldehyde vapors. Each LMOF was exposed to formaldehyde, ethanal, propanal, butanal, pentanal, hexanal, and benzaldehyde vapors, as well as ammonia, ethylamine, diethylamine, trimethylamine, propylamine, butylamine, and aniline vapors. In the case of aldehyde vapors, the same trend was observed in both 1 and 2, with formaldehyde causing the lowest quenching amount with approximately 15% and 10% reduction in emission intensity respectively. The quenching amount increased in trend with molecular weight, and benzaldehyde caused the strongest quenching at about 48% and 80% for 1 and 2, respectively. In all cases, there was no shift in the ligand-centered emission. The higher sensitivity for benzaldehyde in 2 was attributed to its slightly larger channels, which loading simulations indicated could accommodate 2 benzaldehyde molecules per unit cell, unlike the single benzaldehyde per unit cell of 1. Unlike benzaldehyde, the amine vapors induced both emission quenching and red- and blue-shifted emission from both 1 and 2. In each case, aniline caused the strongest emission quenching accompanied by a small redshift, while ammonia induced moderate quenching and a 25 nm blueshift, and ethylamine caused similarly moderate quenching accompanied by a redshift of 40 nm in 1 and 20 nm in 2. This combination of altered emission wavelength and variable quenching efficiency allowed the authors to construct a 2D map of quench % vs. emission shift, potentially permitting the identification of amine vapors through a specific interaction with the sensor LMOFs (Fig. 6).

image file: c7cs00885f-f6.tif
Fig. 6 2D map of quenching efficiency vs. emission wavelength shifts of 1 (left) and 2 (right) when exposed to amine vapors.100 Reproduced from ref. 100 with permission from the American Chemical Society, copyright 2016.

An LMOF displaying strong emission turn-on in the presence of aliphatic amines was recently reported that takes advantage of specific interactions between a hydrogen bond-donating pocket and the low-weight amines methylamine, dimethylamine, and trimethylamine.101 In the two-dimensional strontium-based Sr(H2ABTC)(DMF)(H2O) (H2ABTC = 3,3,5,5-azobenzenetetracarboxylic acid), two of the four carboxyl groups in the ABTC ligand are non-coordinating and remain protonated. Two of these carboxyl groups from neighboring ligands, plus one water molecule coordinated to a neighboring Sr2+, form a hydrogen-rich pocket. Upon introduction of methylamine, dimethylamine, or trimethylamine vapor, hydrogen bonding between this pocket and the amine rigidifies the structure (Fig. 7). Under 10 ppm exposure, this results an approximately two-fold increase in the emission intensity from the LMOF, as well as a redshift in peak emission energy from 558 to 610 nm (Fig. 7). The authors report that the limit of detection is on the order of 10 ppb at room temperature, making this material a promising visual sensor for amine vapors.

image file: c7cs00885f-f7.tif
Fig. 7 (a) Two neighboring 2D sheets of Sr(H2ABTC)(DMF)(H2O), viewed from within the plane of the sheet. (b) Detail of the H-bonding pocket formed between two sheets, with the approximate location of the amine analyte shown. (c) Emission spectra of Sr(H2ABTC)(DMF)(H2O) in the presence of 10 ppm methylamine (MMA), dimethylamine (DMA) and trimethylamine (TMA), as well as nitrogen dioxide, sulfur dioxide, carbon monoxide, and hydrogen gas.101 Reproduced from ref. 101 with permission from the John Wiley & Sons, Inc., copyright 2017.

Another material exhibiting selective emission turn-on in the presence of amine vapors was reported by Mandal et al.102 The material Zn(PA)(BPE) (PA = pamoic acid, bpe = 1,2-bis(4-pyridylethane)) is composed of a three-dimensional framework with 5-fold interpenetration, in which dipole–dipole interactions between amine vapors of ethylenediamine, diisopropylamine, hydrazine, and n-butylamine lead to emission enhancements of 30–100%, with the strength of emission enhancement in trend with the dipole moment of the amine vapor. The material shows no response to a number of other polar and non-polar solvent vapors, and DFT calculations indicate that the higher-lying LUMO of the amines inject electrons into the lower-lying LUMO of Zn(PA)(BPE) upon photoexcitation, resulting in the increased emission intensity from the sensor material.

A turn-on sensor for the widely used aliphatic amide dimethylformamide (DMF) was reported which exhibits an emission enhancement of more than eight fold in the presence of DMF vapor.103 The lanthanide-based LMOF Eu2L3(H2O)4·3DMF (1, L = 2′,5′-bis(methoxymethyl)-[1,1′:4′,1′′-terphenyl]-4,4′′-dicarboxylate) was synthesized solvothermally in DMF and showed strong Eu-centered luminescence. 1 was then soaked in water for three days to give 2, with all remaining DMF molecules exchanged with water. 2 exhibited much weaker luminescence than 1, as the O–H bond vibrations are known to quench emission from Eu3+.103 Upon exposure to a variety of solvent vapors, the luminescence was slightly enhanced (up to 150%), with the enhancement primarily due to the displacement of water molecules within the pore, preventing the quenching of Eu3+ emission. However, upon exposure to DMF vapor, emission intensity was enhanced eight-fold (Fig. 8). The selectivity for DMF was attributed to the fact that the initial solvothermal synthesis of the LMOF took place in DMF. As a result, DMF molecules acted as solvent templates during the synthesis of the material, allowing DMF molecules from the vapor to fit into tailored spaces within the LMOF. These close interactions between the DMF and the LMOF ligands shift the excited state energy in the ligand to increase sensitization of the emissive Eu3+ centers, resulting in turn-on emission.

image file: c7cs00885f-f8.tif
Fig. 8 Emission intensity enhancement of 2 upon incubation in a closed vial with a variety of solvent vapors. MgSO4 was included in one vial as a desiccant, to show that removal of water from 2 resulted in emission turn-on. Reproduced with permission from ref. 103. Reproduced from ref. 103 with permission from the John Wiley & Sons, Inc., copyright 2013.

The detection of aliphatic thiols was reported in 2013, through the use of an Eu3+-loaded LMOF film.43 Films of In-BTC (btc = 1,3,5-benzenetricarboxylate) were prepared solvothermally on a Si wafer that was coated with Pt nanoparticles, which served to increase the roughness of the surface and improve MOF nucleation. The resulting MOF possessed MIL-100-type connectivity, with two of the BTC carboxylates coordinating with In2+ while the third projected uncoordinated into the pore. Eu3+ was postsynthetically added to the material at a ratio of Eu[thin space (1/6-em)]:[thin space (1/6-em)]In = 0.071 to create In-BTC⊃Eu, with the added Eu3+ coordinated to the free carboxylate group. The resulting In-BTC⊃Eu thin film showed characteristic Eu3+ emission with peak emission intensity at 618 nm. The excitation spectrum at 618 nm emission showed a broad band between 200–300 nm corresponding to absorbance by the btc ligand, demonstrating that the lanthanide was efficiently sensitized. Upon exposure to saturated vapors of 1-butanethiol and 1,2-ethanedithiol, emission was quenched by 94% within 20 seconds and 92% within 120 seconds for each analyte respectively. The authors claim that the emissive Eu3+ within the LMOF pores was coordinatively unsaturated, allowing the thiols to bind with open sites on the metal and quench emission.

A porous Cu(I)-based LMOF (H2O⊂Cu2(L)2I2, L = 1-benzimidazolyl-3,5-bis(4-pyridyl)benzene) able to detect the presence of small halocarbons on the ppm level through an emission turn-on was reported by Dong et al.55 When exposed to CHCl3 and CH2Cl2, emission from the material turned on, with a two-fold increase in intensity and a slight blueshift. For both analyte molecules, the authors report a limit of detection of 10 ppm. In order to identify the mechanism of this emission enhancement, H2O⊂Cu2(L)2I2 was incubated in the analyte solvents, and single crystal analysis was able to locate the adsorbed solvent molecules within the pore. Instead of coordinating with the metal ions, CHCl3 and CH2Cl2 were found to form weak hydrogen-bonding interactions with the ligand molecule. The analyte-loaded sample was also found to have a slightly blueshifted absorbance band. The authors propose that the hydrogen bonds serve to rigidify the LMOF structure and alter ligand energy levels, making the emission more efficient.

A LMOF was constructed by linking a luminescent dodecanuclear silver chalcogenide cluster [(Ag12(StBu)6(CF3COO)6(CH3CN)6)]·CH3CN (Ag12) into a framework with bpy (4,4′-bipyridine) ligands replacing the previously coordinating CH3CN groups.54 The resulting LMOF Ag12bpy (Ag12(StBu)8(CF3COO)4(bpy)4) gave strong green emission at 500 nm in the absence of O2, but in the presence of O2, emission was strongly quenched. However, when exposed to a group of VOC solvent vapors including chloroform, emission from Ag12bpy was turned on again through a fast displacement mechanism with response in under one second, as VOC molecules replaced O2 within the material's pores (Fig. 9b). Additionally, the position of the emission peak redshifted by up to 50 nm based on the identity of the VOC adsorbed, with chloroform inducing a redshift of approximately 25 nm (Fig. 9a), thus allowing the identification of the VOC vapor. Both adsorption data and single crystal analysis confirmed that an average of four chloroform molecules were present per unit cell.

image file: c7cs00885f-f9.tif
Fig. 9 (a) Shift in emission from Ag12bpy when exposed to various VOC vapors. (b) Emission turn-on in Ag12bpy when exposed to increasing concentration of CHCl3 vapors.54 Reproduced from ref. 54 with permission from the Nature Publishing Group, copyright 2017.

2.3. Toxic and hazardous gas detection

Pollution resulting from combustion exhaust is the largest source of exposure to hazardous gases, include CO, NO, NO2, and SO2. These compounds are acutely toxic, and in the case of NO2 and SO2, are contributors to acid rain. Other toxic gases, like HCl, can be produced by burning plastics or other polymers. Ammonia vapor is among the most common indoor pollutants, and can be released from cleaning products or outgas from building materials.106 Detecting these and other harmful gases and vapors is an important aspect of air quality monitoring; however, when compared with aromatic and aliphatic VOCs, relatively fewer LMOFs have been reported for the detection of these species. While ammonia detection is fairly well developed, the LMOF-based detection of many other common hazardous gases including SO2 or CO have not been reported, to the best of our knowledge.

A LMOF based on UiO-66 (Zr6O4(OH)4(bdc)6, bdc = benzene-1,4-dicarboxylate) able to detect the toxic gases NO, NO2, and Br2 was reported in 2015 by Kaskel et al. which targets these three gases by their ability to act as strong oxidizers.107 By postsynthetic modification, dihydro-1,2,4,5-tetrazine-3,6-dicarboxylate (dhtz) ligand molecules replaced bdc ligands within the structure of UiO-66 at approximate a 5[thin space (1/6-em)]:[thin space (1/6-em)]1 bdc[thin space (1/6-em)]:[thin space (1/6-em)]dhtz ratio to give UiO-66(dhtz). This dihydro-1,2,4,5-tetrazine unit was chosen for it's ability to be easily oxidized to 1,2,4,5-tetrazine (tz), and for the fact that this oxidation results in a clear colorimetric shift from yellow to pink as a result of increased blue-green light absorbance. Upon exposure to NO, NO2, and Br2 gas, the dhtz ligands within UiO-66(dhtz) were oxidized to tz, giving UiO-66(tz) and a clear color change from yellow to pink (Fig. 10). This oxidation could be fully reversed by suspending UiO-66(tz) in an aqueous solution with the reducing agent sodium dithionite, allowing the material to be used again. PXRD and gas adsorption measurements demonstrated that the structure's crystallinity and porosity was not affected by the postsynthetic ligand replacement, oxidative sensing, or reductive regeneration.

image file: c7cs00885f-f10.tif
Fig. 10 (top) Postsynthetic ligand replacement of bdc with dhtz to give UiO-66(dhtz). (bottom) The fully-reversible oxidation of UiO-66(dhtz) to UiO-66(tz) in the presence of NO, NO2, or Br2 gases.107 Reproduced from ref. 107 with permission from the Royal Society of Chemistry, copyright 2015.

A nanocrystalline, lanthanide-based LMOF has been reported for the sensitive, selective detection of ammonia vapor with the detection giving both a photoluminescent and a colorimetric signal.108 Nanocrystalline Ga(OH)bpydc (2,2′-bipyridine-4,4′-dicarboxylic acid) was prepared solvothermally, with Eu3+ emission centers added postsynthetically. These Eu3+ ions coordinate with the bipyridine moiety of the ligand, resulting in efficient sensitization and Eu-centered emission from the LMOF. When exposed to a number of indoor air pollutants including benzene and formaldehyde vapors, there is very slight emission enhancement or quenching—on the order of ±5–10%. However, when exposed to NH3 a 76% quench of Eu-based emission was observed, coupled with strong enhancement of ligand-based emission. This lead to a clear spectral and colorimetric change with the extremely low limit of detection of is 2.4 ppm NH3, which is well below the 50 ppm workplace limit.108 The response was also rapid, with the first change by 30 seconds and a complete response after 240 seconds. Furthermore, the material can be regenerated and reused by simple exposing the material to ambient air (Fig. 11). As the PXRD and Eu-centered emission lifetime of the LMOF is unchanged, the structure is stable in the presence of the analyte, and the analyte does not interact directly with the Eu3+ ion. Instead, the analyte interacts with the ligand. IR peaks of ligand skeleton shift when exposed to analyte, and ligand absorbance increased, along with a 24 nm redshift in ligand emission, indicating that H-bonding between the ligand and analyte lowered the ligand π* orbitals. This prevented effective sensitization of the Eu, and resulted in the observed signal.

image file: c7cs00885f-f11.tif
Fig. 11 (a) Intensity of emission from Eu@Ga(OH)bpydc under different concentrations of NH3 vapor. (b) Emission spectra of Eu@Ga(OH)bpydc following exposure to NH3 vapors at various exposure times, with the emission intensity at 614 nm vs. time inset. (c) Emission spectra of Eu@Ga(OH)bpydc following exposure to ambient air, demonstrating the recovery of luminescence, with the peak at 614 nm inset. (d) Intensity of the 614 nm emission from Eu@Ga(OH)bpydc before (left) and after (right) exposure to NH3 vapors over three cycles.108 Reproduced from ref. 108 with permission from the Royal Society of Chemistry, copyright 2016.

Ammonia was also detected selectively at high temperatures through interactions with Zn2+ and Mg2+ open metal sites (OMS) in Zn2(tcpe) (tcpe = tetrakis(4-carboxyphenyl)ethylene) and Mg(H2DHBDC) (H2DHBDC2− = 2,5-dihydroxybenzene-1,4-dicarboxylate), respectively, that induced a significant redshift in emission from the LMOFs.109 The behavior was first noted in Zn2(tcpe), which forms paddlewheel SBUs with four ligand carboxylates forming the paddles and two water molecules coordinating in the axial positions. At room temperature, emission from the material redshifts when exposed to NH3, triethylamine, and ethylene diamine, but once heated to 100 °C, redshifted emission is only observed when the material is exposed to NH3. DFT calculations demonstrated that this was likely because of the interactions between NH3 and the OMS on Zn following removal of the coordinating water are stronger than interactions between other analytes and the Zn2+ OMS. However, this interaction was also accompanied by an irreversible structure change as observed in PXRD. This led the authors to consider Mg(H2DHBDC), as it likewise possessed coordinating solvent that could be removed under heating and maintained emission intensity at elevated temperatures. Upon exposure to NH3 vapor at 100 °C, redshifted emission was again observed from the MOF. However, as the interaction between NH3 and Mg2+ isn’t as strong as the interaction between NH3 and Zn2+, the coordinated ammonia could be removed by evacuation for 15 minutes, allowing the sensor material to be reused.

A Eu3+-based 1D coordination polymer that stacks to form a 3D porous network was used to detect gaseous HCl through the protonation of a non-coordinative basic site on the ligand.52 In the EuH(L)2(NO3)2 (EuL, L = 2-(2-pyridin-2-yl)quinoline-4-carboxylic acid) is composed of an infinite PBU, with each Eu3+ center coordinated to carboxylate groups from four ligand molecules plus two NO3 ions, with each carboxylate group bridging two Eu3+ ions to form a 1D chain. EuL emits characteristic Eu3+-centered emission with efficient sensitization by the ligand. However, in the presence of HCl gas, the free pyridine moieties are protonated, and emission from the material is quenched. Time-dependent Hartree–Fock calculations were performed to determine the energy levels of the singlet and triplet excited states in the protonated and non-protonated ligand, in order to compare them to the energy level of the Eu3+ and assess how protonation might change the emission mechanism. The authors found that the non-protonated ligand triplet state was located 3500 cm−1 higher than the 5D0 transition of Eu3+, which is in the optimal range for energy transfer from the ligand to the metal. However, upon protonation, the energy of the ligand triplet state drops to just 900 cm−1 above the Eu3+ 5D0 transition, allowing back-transfer from the Eu3+ to the ligand and quenching emission.

The fast and sensitive detection of HCl gas through the use of a copper(I) iodide-based MOF Cu4I4L (1, L = 5,5′,5′′-(2,4,6-triethylbenzene-1,3,5-triyl)tris(2-(pyridin-4-yl)-1,3,4-oxadiazole)) was recently reported by Dong et al.1101 was synthesized under ambient conditions through the combination of CuI and L in acetonitrile, and gave a structure with two crystallographically-distinct Cu4I4 clusters linked into a doubly interpenetrated 3D framework by the ligand L. Upon exposure to 200 ppm HCl, 1 changed color from orange to dark brown (Fig. 12). While PXRD identified no structural change, a starch assay identified molecular iodine present in the pores of the material, with Raman analysis, XPS, and ion chromatography confirming that exposure to the HCl gas induced an I/Cl ion exchange in the Cu4I4 core that liberated I. The I was then oxidized by atmospheric O2 in the presence of H+ to give I2, which in turn provided the colorimetric shift observed (Fig. 12). The colorimetric shift observed was extremely sensitive, with HCL concentrations of as little as 4 ppb causing a clear difference in color. Because of the ion-exchange nature of the interaction, 1 is extremely selective for HCl gas, and no response was observed in the presence of other similar gaseous acids, such as HF, HBr, HI, HOAc, HNO3, and HCLO4. To increase the practical applicability, a composite material of 1 embedded in a polymer matrix was prepared in a single-step, one-pot process. CuI and L were combined under ambient conditions in a DMF/acetonitrile solution containing the polymer binder polyvinylidene fluoride. Acetonitrile was then removed from the solution under vacuum, resulting in a homogenous suspension of micro-sized 1 and polymer binder in DMF, with the content of 1 being up to 69 wt%. The suspension was cast and dried at 90 °C to give a mixed-membrane matrix (1@MMM). 1@MMM was homogenous and maintained the porosity of 1. When exposed to HCl vapor, 1@MMM showed the same sensitivity with a visual detection limit of 3.2 ppb and a luminescent detection limit of 1.6 ppb; both are significantly lower than the 5 ppm workplace exposure limit. Additionally, because of the film's relative thinness and correspondingly increased contact area, the response of 1@MMM is much faster, with full color change observable after 1 minute of exposure.

image file: c7cs00885f-f12.tif
Fig. 12 (a) Sensing mechanism of 1 upon exposure to HCl gas. (b) Samples of 1 exposed to various concentrations of HCl gas. (c) Emission from 1 under 370 nm excitation following exposure to various concentrations of HCl gas. (d) Samples of 1@MMM following exposure to various concentrations of HCl gas. (e) SEM images of 1@MMM following exposure to HCl gas at high (top left) and low (top right) magnification, with EDS mapping shown for I, F, and Cl. (f) CO2 adsorption isotherms for 1 (black) and 1@MMM (red). (g) Pore width distribution of 1 and 1@MMM.110 Reproduced from ref. 110 with permission from the Royal Society of Chemistry, copyright 2016.

2.4. Detection of CWAs

Unlike the previously described chemical species, whose toxicity arises incidentally, chemical warfare agents (CWAs) have been designed to maximize the negative impact they can have on human health. The two main categories of modern CWAs are blistering agents, such as sulfur mustard or lewisite, and nerve agents, such as tabun, sarin, and VX. MOFs have been reported for detoxification of chemical warfare agents and liquid-phase absorbance and luminescent sensing, but vapor-phase detection has been lacking.111–115 This is a concern, as these weaponized nerve agents are typically dispersed in the vapor-phase. Because of their acute toxicity, designing and testing sensor materials for nerve agents poses a practical challenge, so less toxic analogues are often used. Reports of LMOFs able to detect CWAs in the vapor phase are very limited at this time. However, those that have been reported are very recent, indicating that this important field may be beginning to grow.

A weakly-emissive 1D coordination polymer that is converted into a strongly luminescent 3D framework upon exposure to the thioethers dimethyl sulfide (DMS) and diethyl sulfide (DES)—an analogue for the blistering agent di-(2-chloroethyl) sulfide, or sulfur mustard—in the vapor phase was recently reported by Leznoff et al.116 The coordination polymer Cu1/2Au1/2CN was initially chosen because of the strong affinity of Cu open metal sites for thioethers, and the author's previous work preparing the LMOFs [Cu1/2Au1/2CN]2(DMS) and [Cu1/2Au1/2CN]2(DES) in solution.117 However, the authors found that Cu1/2Au1/2CN was unresponsive to DMS and DES vapors, as strong inter-chain Au–Au interactions enforced structural rigidity that prevented the necessary rearrangement imposed by the coordination of the analyte with the Cu(I) centers. The amount of Au(I) within the material was therefore reduced to give Cu2/3Au1/3CN, in order to limit the amount of direct Au–Au interactions. Upon exposure to DMS and DES vapors, emission from the material was strongly turned on (Fig. 13) and shifted from 380 nm under UV excitation to 460 nm (DMS) and 420 nm (DES) (Fig. 13). PXRD analysis confirmed that the material, following exposure to the thioether vapors, was isostructural with the previously reported [Cu1/2Au1/2CN]2(DMS) and [Cu1/2Au1/2CN]2(DES), confirming that the thioether sulfur was bonded to the Cu(I) atoms. The sample could be regenerated by either blowing thioether-free air over the sample for extended time, or by heating at 120 °C for 15 minutes. Following regeneration, thioether sensing could be repeated for multiple cycles (Fig. 13).

image file: c7cs00885f-f13.tif
Fig. 13 (a) Cu2/3Au1/3CN before and after exposure to DMS vapor, with the response to DES vapor being visually identical. (b) Excitation and emission spectra of Cu2/3Au1/3CN before and after exposure to DMS or DES. (c) Intensity of emission at 418 nm from Cu2/3Au1/3CN over multiple cycles of exposure and regeneration, with regeneration achieved by heating the sample at 120 °C for 15 minutes.116 Reproduced from ref. 116 with permission from the Royal Society of Chemistry, copyright 2017.

The MOF HKUST-1 (Cu3(btc)2, btc = 1,3,5-benzenetricarboxylate) was recently reported to detect dimethyl chlorophosphate (DMCP), an analogue for G-series nerve agents including sarin and VX, in the vapor phase through a clear colorimetric shift.118 HKUST-1 was chosen for its open-copper sites, which the authors hoped with increase adsorption of the DMCP through coordination with the phosphate oxygen. HKUST-1 was mounted on a cotton textile (T-M) and exposed to DCMP vapor, resulting in a clear color change from turquoise to yellow (Fig. 14). Samples of HKUST-1 in combination with graphitic carbon nitride (MOFgCN) and oxidized graphitic carbon nitride (MOFgCNox) were also prepared and mounted on a cotton textile (T-MG and T-MGox, respectively). These composite materials not only showed colorimetric shifts when exposed to DCMP vapor, but T-MGox showed impressive DCMP absorbance of 670 wt% relative to Cu(I) and was also capable of degrading DCMP through visible light-driven hydrolysis.

image file: c7cs00885f-f14.tif
Fig. 14 Color changes of cotton textile, T-M, T-MG, and T-MGox when exposed to DCMP vapors over time.118 Reproduced from ref. 118 with permission from the Royal Society of Chemistry, copyright 2017.

3. Adsorptive capture of toxic and hazardous gases and vapors by MOFs

In addition to effective detection capture and sequestration of toxic and hazardous species is also vital to prevent their potential threat on human health and/or the environment. Adsorptive capture represents a promising technology for the removal of toxic and hazardous gases and vapors as it is energy efficient and environmentally friendly. Various adsorbent materials have been proposed for this process.119–121 For example, carbon-based broad-spectrum filter is designed for removing toxic industrial gases. However, it suffers from low selectivity as amorphous carbons have broad pore size distribution.122 Additionally, carbon-based adsorbents usually have relatively weak affinity to polar adsorbates such as NH3 or H2S, resulting in low capacity.122 Thus, tailor-made adsorbent materials are needed for efficient removal of target species under specific conditions. The highly tunable nature of MOFs with respect to their composition, porosity, pore structure, and surface functionality renders them enormous advantages for adsorptive capture of toxic and hazardous gases. Though in some aspects the exploration is still in its early stage, MOFs have shown great potential and are promising for implementation in real-world systems.

3.1. Adsorption of fluorocarbons and chlorofluorocarbons

Fluorocarbons (FCs, including hydrofluorocarbons (HFCs)), and chlorofluorocarbons (CFCs, including hydrochlorofluorocarbons (HCFCs)) refer to hydrocarbon derivatives where one or more hydrogen atoms in hydrocarbons have been replaced by fluorine and/or chlorine atoms. They represent an important category of organic compounds that are ubiquitously involved in industry and our daily life. Their applications include use as solvents, refrigerants, and anesthetics, to name a few.39,123 However, they have also given rise to tremendous safety and environmental concerns. Many FCs are greenhouses species with 100 year global warming potentials (GWPs, Table 1) 3–4 orders of magnitude higher than carbon dioxide, though they are present at a lower concentration in air.124 CFCs, like the well-known Freon, are active ozone reducers which have destructive effects on the ozone layer. In light of their significant environmental impact, the use of CFCs has been heavily regulated since the late 1970s. For example, the United States banned the use of CFCs such as Freon in aerosol cans in 1978. In addition, a series of international treaties, including the well-known Montreal Protocol agreed to in 1987, have been ratified to protect the ozone layer by phasing out the production of substances that are responsible for ozone depletion. Since the beginning of the regulations against ozone depleting species, and especially since the adoption of the Montreal Protocol, the atmospheric concentration of many CFCs and related chlorinated hydrocarbons has noticeably decreased. But despite the continued regulatory actions taken against the use of CFCs, they have never been completely phased out and still pose a threat to both the ozone layer and the climate. This is because the use of CFC-producing products has never been completely banned globally and related equipment is still in use, particularly in some developing countries. In addition, while the production and consumption of CFCs are regulated, emissions from products that already contain CFCs are unregulated. These products—refrigerators, air conditioners, and others—are a constant source of CFC emissions. More importantly, as the interim replacements for CFCs, HCFCs can also deplete the ozone layer, though to a lesser extent. In this context, the capture and sequestration of FCs and CFCs are imperative to expedite the complete elimination of ozone depleting species. FCs/CFCs adsorption related applications also include separation and/or enrichment of FCs/CFCs for recycling, FCs/CFCs-based adsorption heat pumps and so on.
Table 1 Selected FCs and CFCs with their boiling point and 100 year GWP (global warming potential)
Common name Formula Boiling point (°C) 100 year GWP (vs. CO2)
CFC-11 CCl3F 23.77 4660
CFC-12 CCl2F2 −29.8 10[thin space (1/6-em)]200
CFC-13 CClF3 −81 13[thin space (1/6-em)]900
CFC-113 CCl2FCClF2 47.7 5820
CFC-114 CClF2CClF2 3.8 8590
CFC-115 CClF2CF3 −38 7670
HFC-23 CHF3 −82.1 12[thin space (1/6-em)]400
HFC-32 CH2F2 −52 677
HFC-41 CH3F −78.4 116
HFC-125 CHF2CF3 −48.5 3170
HFC-134 CHF2CHF2 −23 1120
HFC-134a CH2FCF3 −26.3 1300
HFC-143 CH2FCHF2 5 328
HFC-143a CH3CF3 −47.6 4800
HFC-152 CH2FCH2F 31 16
HFC-152a CH3CHF2 −25 138
HFC-161 CH3CH2F −37.7 4
HFC-227ea CF3CHFCF3 −16.4 3350
HFC-236fa CF3CH2CF3 −1.4 8060
HFC-245fa CHF2CH2CF3 15.3 858
HCFC-22 CHClF2 −40.8 1760
HCFC-123 CHCl2CF3 27.82 79
HCFC-124 CHFClCF3 −12 527
HCFC-141b CCl2FCH3 32 782
HCFC-142b CClF2CH3 −9.2 1980
HCFC-225ca CF3CF2CHCl2 51 127
HCFC-225cb CClF2CF2CHClF 56 525
PFC-14 CF4 −127.8 6630
PFC-116 C2F6 −78.2 11[thin space (1/6-em)]100
PFC-218 C3F8 −36.7 8900
PFC-31-10 C4F10 −1.7 9200
PFC-41-12 C5F12 28 8550
PFC-51-14 C6F14 56 7910
Halon-1211 CBrClF2 −3.7 1750

Thallapally and co-workers125 explored the adsorption of a series of fluorocarbon derivative refrigerants in selected MOFs including NiDOBDC, CoDOBDC, MIL-101-Cr, and MIL-100-Fe. MDOBDC (M = Ni and Co) features a microporous framework with a high density of open metal sites (OMSs) while MIL-101-Cr and MIL-100-Fe possess hierarchical pore structures with mesoporous cages connected through microporous windows. Thus the selected MOFs allow the effect of pore morphology and functionality on refrigerant adsorption to be investigated. The studied refrigerants include R-12 (CFC-12, CF2Cl2), R-13 (CFC-13, CClF3), R-14 (PFC-14, CF4), R-22 (HCFC-22, CHClF2), and R-32 (HFC-32, CH2F2). All adsorbents show typical Type I adsorption profiles at room temperature (Fig. 15). Adsorption of R-12 on NiDOBDC and CoDOBDC reaches saturation at relatively low pressure (P/P0 = 0.01) with an uptake amount of 4.58 mol g−1, which is more than twice the capacity of MIL-101-Cr (<2 mmol g−1) at the same pressure. However, the saturation capacity of MIL-101 (at P/P0 = 0.6) reaches 15 mmol g−1, which is much higher than that of the other compounds. Grand Canonical Monte Carlo (GCMC) simulations suggest that at low pressure the primary adsorption site of R-12 is the OMSs for MDOBDC, while for MIL-101 it is preferentially adsorbed in the small pockets. By performing a column breakthrough measurement with a gas mixture of 90% He, 2% R-22 and 8% of R-12, the authors found MIL-101 can effectively separate the mixture into individual fractions. In light of the high adsorption capacity, especially at low pressure, and the separation ability toward FCs and CFCs, the authors concluded that these MOFs may be promising for FCs/CFCs adsorption/separation related applications.

image file: c7cs00885f-f15.tif
Fig. 15 (a) Molecular structure of FCs and CFCs. (b) Adsorption isotherms of R-12 at 298 K in MDOBDC (M = Ni, Co), MIL-100-Fe, and MIL-101. (c) Adsorption isotherms of various refrigerants in MIL-101 at 298 K. (d) Experimental breakthrough of adsorption bed packed with MIL-101, a mixture of 90% He, 2% R-22, and 8% R-12 fed through the column with a flow rate of 0.25 mL min−1.125 Reproduced from ref. 125 with permission from the Nature Publishing Group, copyright 2014.

Miljanic and co-workers126 reported the adsorption of FCs and CFCs by a noncovalent organic framework built on a highly fluorinated trispyrazole-based organic molecule (Fig. 16a and b). The framework exhibits high thermal stability (stable up to 250 °C), chemical stability (stable in common organic solvents, water, acidic and basic aqueous solutions) and high porosity (BET surface area: 1159 m2 g−1). As expected, the material shows hydrophobic behavior with a contact angle of 132 ± 1° and it does not take up water even at 90% relative humidity, but exhibits favorable adsorption toward FCs and CFCs. It adsorbs 74 wt% of perfluorohexane at room temperature and no loss of capacity was observed after 20 adsorption/desorption cycles. In addition, the adsorption is very fast and reaches equilibrium in ∼20 seconds. The compound shows similar adsorption behavior toward other FCs and CFCs including chloroform, dichloromethane, CFC-113 (CCl2FCClF2), and HCFC-225ca (CF3CF2CHCl2). In a follow-up work,127 the authors modify the organic molecule to a tritetrazole ligand and subsequently incorporate it into a Cu-based metal–organic framework (MOFF-5, Fig. 16c). It shows a BET surface area of 2445 m2 g−1 with both micropores and mesopores. Adsorption of a number of FCs and CFCs was tested on MOFF-5, and high uptake capacity was observed for most of them due to its fluorinated structure and high porosity. MOFF-5 adsorbs as much as 225 wt% of perfluorohexane at room temperature, and the adsorption is complete within seconds of exposure (Fig. 16d).

image file: c7cs00885f-f16.tif
Fig. 16 (a) Molecular structure of the fluorinated monomer and (b) crystal structure of the noncovalent organic framework. (c) Crystal structure of MOFF-5 and (d) its adsorption of perfluorohexane (C6F14) at room temperature. Black lines indicate the parts of the program when MOFF-5 was exposed only to N2 stream, while red line describes the section of the program when N2 carrying C6F14 vapor was passed over MOFF-5.126,127 Reproduced from ref. 126 and 127 with permission from the Nature Publishing Group and the John Wiley & Sons, Inc., copyright 2014 and 2015, respectively.

Several MOF materials have been evaluated for the adsorption of R-22 (HCFC-22, CHClF2). R-22 is a commonly used refrigerant. Due to its relatively low ozone depletion potential, R-22 was selected as an alternative to the highly ozone-depleting R-11 and R-12. However, the use of R-22 is no longer considered acceptable as a result of the enforcement of regulations against ozone depleting species. It has been mostly phased out in the United States and European Union, but its use in developing countries continues to increase owing to its high demand. Chen et al.128 reported the adsorption of R-22 by a series of isoreticular MOFs with a formula of Zn4O(bpz)2(ldc) (bpz = 3,3′,5,5′-tetramethyl-4,4′-bipyrazlate and ldc = linear dicarboxylates including 1,4-benzenedicarboxylate, naphthalene-1,4-dicarboxylate, and biphenyl-4,4′-dicarboxylate). The porosity and pore size/shape of these compounds can be systematically tuned by changing the linear dicarboxylate linkers. R-22 adsorption isotherms were collected on these compounds at 273 and 313 K, and they all exhibit high adsorption capacity (80–120 wt% at 273 K and 1 bar, 65–75 wt% at 313 K and 1 bar). The high working capacity, relatively large adsorption enthalpies, and fast diffusion make these compounds promising candidates for R-22 capture and heat transformation systems.

By a post-synthetic variable-spacer installation (SVSI) strategy, Su and co-workers129 showed the fine-tuning of the porosity and pore surface of a series of Zr-MOFs modified from a prototype Zr-MOF, Zr6O8(H2O)8(L1)4 (LIFM-28, L1 = 2,2′-bis(trifluoromethyl)-4,4′-biphenyldicarboxylate). These materials have BET surface areas ranging from 940 to 1588 m2 g−1. Interestingly, it was found that the MOFs modified by ligand insertion show much higher adsorption capacity for R-22 (70–130 cc g−1 at 298 K and 1 bar) and excellent R-22/N2 selectivity (170–290 calculated by IAST). In a follow-up work,65 the authors applied the pore engineering strategy to PCN-700, an isostructural MOF of LIFM-28. By post-synthetic insertion of different dicarboxylate linkers, the authors were able to tune the pore environment and the adsorption performance of the materials. These compounds are highly porous, with BET surface areas of 1496–2222 m2 g−1, and they show both high adsorption capacity and selectivity toward R-22 (Fig. 17). The R-22 uptake capacity and R-22/N2 IAST selectivity are 150–220 cc g−1 (273 K and 1 bar) and 600–1000 (273 K and 1 bar, R-22/N2 = 10[thin space (1/6-em)]:[thin space (1/6-em)]90), respectively. These values are significantly higher than that of the pristine PCN-700. This was attributed to the pore space partition and pore surface modification by the inserted linkers.

image file: c7cs00885f-f17.tif
Fig. 17 R-22 adsorption isotherms of LIFM-28np, LIFM-29, LIFM-30, LIFM-31, LIFM-32, and LIFM-33 at (a) 273 K and (b) 298 K. (c) Adsorption isotherms of R-22 and N2 on LIFM-32 at 273 K and 298 K. (d) IAST selectivity of R-22/N2 on LIFM-32at 273 K and 298 K.65 Reproduced from ref. 65 with permission from the Royal Society of Chemistry, copyright 2017.

Very recently, Motkuri and co-workers130 reported the adsorption of another widely used refrigerant R-134a (HFC-134a, CH2FCF3) by Ni-MOF-74 and its derivative compounds. Through ligand modifications, the authors was able to engineer the pore structure and functionality of the MOFs. The saturated uptake capacities of R-134a are 0.58, 0.75, and 0.77 g g−1 for Ni-MOF-74, Ni-MOF-74-BPP (BPP = 3,3′-dioxido-4,4′-biphenyldicarboxylate, biphenyl with para-COOH), and Ni-MOF-74-TPP (TPP = 3,3′-dioxido-4,4′-triphenyldicarboxylate, triphenyl with para-COOH), respectively, at 298 K while Ni-MOF-74 shows the highest adsorption enthalpy of 50.6 kJ mol−1. In situ FTIR analysis indicates the primary adsorption site for R-134a is the open Ni2+ centers.

3.2. Capture of radioactive gases and vapors

Among all energy sources which serve as alternatives to carbon-based fossil fuels, nuclear energy represents one of the most promising candidates in view of its low cost, high energy density, and low emission of greenhouse gases. Nuclear energy currently provides 11% of the world's electricity, and its contribution will undoubtedly continue to increase in light of the rapidly growing global energy demand.36,131 However, in the process of mass implementation of nuclear power, we must safely capture and sequester the associated radioactive nuclear waste. While tremendous effort has been made to reprocess and recover heavy radioactive elements such as uranium and plutonium, less attention was paid to the volatile radioactive species, which include iodine, organic iodides, and krypton among others. These species are radiotoxic and highly volatile and must therefore be captured and removed from the off-gas mixtures to prevent their release into the environment.

In the process of nuclear waste management, off-gas streams containing volatile radioactive species (129I2, CH3129I, CH3CH2129I, 14CO2, 85Kr etc.) as well as H2O, nitric acid vapor, and NOx are produced by dissolving used fuel rods in hot, concentrated (3–5 M) nitric acid. Thus, the proposed capture process involves selective adsorption of the target species from thsee off-gas mixtures.132 The capture usually relies on chemisorption or strong physisorption, as the interaction must be specific and selective for the targeted species in the presence of non-radioactive off-gas components. The current technology for I2 removal involves the use of silver exchanged zeolites which convert I2 to AgI or AgIO3.133 However, these Ag-based sorbents have several disadvantages, including high cost, poor recyclability, and low capture capacity. In this context, various sorbent materials (silica, alumina, zeolites, activated carbons etc.) have been investigated for this application. A very recent example involves the use of all-silica zeolites for the capture of iodine and organic iodides.132 The authors found that hydrophobicity-intensified silicalite-1 (HISL), an exceptionally hydrophobic sorbent that is stable under highly acidic conditions, can effectively adsorb I2, CH3I, and CH3CH2I from a simulated acidic off-gas stream containing HNO3 vapor and its decomposed products. At room temperature, its I2 capture capacities under dry and simulated off-gas conditions are 53 wt% and 30 wt%, respectively. Similar capacities were observed for CH3I and CH3CH2I. This compound outperforms many other sorbent materials under similar experimental conditions, and especially under simulated off-gas conditions. However, its disadvantages include relatively low adsorption capacity and poor capture capability at increased temperature. The former is limited by the low porosity of the sorbent, while the latter is attributed to the relatively weak adsorption affinity.

Metal–organic frameworks feature high porosity and exceptional tunability which render them enormous advantages in addressing this challenge. Over the past few years, MOF materials have exhibited superior performance in the capture of volatile radioactive species.

3.2.1. Iodine capture. Among all volatile radionuclides, iodine poses exceptional issues because of the particularly long half-life of 129I (1.57 × 107 years). In the early studies of I2 adsorption by MOFs, I2 was selected as a probe molecule to investigate the guest inclusion and removal behaviors. For example, Zeng and co-workers134 reported a zinc based rigid-pillared MOF, Zn3(DL-lac)2(pybz)2 (DL-lac = DL-lactic acid, pybz = 4-(pyridine-4-yl)benzoic acid), and its controlled uptake and release of iodine. By suspending the desolvated MOF crystals in a cyclohexane solution of iodine, the authors observed a visual color change of the crystals from colorless through yellow and dark brown to black with time (Fig. 18). The controllable release of iodine from the I2-loaded MOF into organic solvent was also investigated. However, in these early explorations, the adsorption capacities were relatively low, since the sorbent materials were not designed for I2 capture. Additionally, the I2 loading experiments were performed at room temperature and commonly in solutions that are not actually relevant to nuclear waste management, which involves capture of I2 vapor at relatively high temperature (75 °C). The first detailed study of iodine vapor capture was reported by Nenoff and coworkers.135 In this study, ZIF-8 was selected for I2 adsorption because of its large surface area (1810 m2 g−1), suitable pore aperture (3.4 Å), and high thermal and chemical stability.136 The iodine loading was performed under typical fuel reprocessing conditions, ca. 350 K and ambient pressure. The maximum adsorption capacity of I2 on ZIF-8 was observed to be 125 wt% and was reached in several hours. This uptake amount was much higher than that of the traditional zeolite materials. It was observed that most of the adsorbed I2 molecules were strongly trapped in the pores of ZIF-8 (Fig. 19a); the weakly surface-adsorbed I2 (25 wt%) were removed by heating the sample at 400 K for 1 hour, but no additional release of I2 molecules was detected before the structure collapsed at 575 K. The synchrotron powder X-ray diffraction (PXRD) analysis revealed that the framework integrity was well maintained for I2 loadings up to 70 wt% (1.3 I/Zn), beyond which the material lost its crystallinity. Following the loss of long-range order, however, the cage connectivity was actually retained, as indicated by PDF (pair distribution function) analysis. To evaluate the processability of ZIF-8 for I2 capture from nuclear waste, the authors performed additional I2 adsorption experiments on the extruded pellet form that is typically employed in real-world separation processes, and observed no change in I2 adsorption performance compared to its powder form. In a follow-up study,137 the authors found the retention of adsorbed I2 molecules in ZIF-8 could be further improved by pressure-induced amorphization. Under a pressure of 0.34 GPa, the I2 loaded ZIF-8 structure was amorphized, whereas the local cage structure with the captive I2 remains intact. The mass losses corresponding to I2 desorption shifted to higher temperatures by up to 150 °C for amorphized materials.
image file: c7cs00885f-f18.tif
Fig. 18 (a) Crystal structure of Zn3(DL-lac)2(pybz)2 showing the channels and (b) sketch of I2 molecules in the channels. (c) Photographs showing the visual color change when a single crystal was immersed in the cyclohexane solution of I2 (0.1 M L−1). (d) I2 enrichment progress when 100 mg of crystals were soaked in 3 mL of a cyclohexane solution of I2 (0.1 M L−1).134 Reproduced from ref. 134 with permission from the American Chemical Society, copyright 2010.

image file: c7cs00885f-f19.tif
Fig. 19 (a) Two I2 adsorption sites in the cage of ZIF-8. (b) I2 adsorbed in the small (left) and big (right) cage of HKUST-1.135,138 Reproduced from ref. 135 and 138 with permission from the American Chemical Society, copyright 2011 and 2013, respectively.

Cu-BTC (or HKUST-1, BTC = 1,3,5-benzenetricarboxylates) is another commercially available prototype MOF with high porosity.139 Nenoff and co-workers138 investigated I2 capture by Cu-BTC and the competitive adsorption of water under humid conditions. Since the off-gas streams of nuclear waste are humid, it is crucial that the proposed sorbent materials retain their capture performance in the presence of moisture. In this study, the iodine adsorption experiments were carried out at 75 °C with a 1[thin space (1/6-em)]:[thin space (1/6-em)]1 ratio of I2[thin space (1/6-em)]:[thin space (1/6-em)]H2O vapor approximating the condition for real-world nuclear fuel reprocessing. During the adsorption process, the adsorbed amount was measured gravimetrically and confirmed by micro X-ray fluorescence. The maximum I2 uptake was determined to be 175 wt%, corresponding to a loading of 3 I/Cu. This uptake capacity surpassed that of ZIF-8, the previous record holder. Unlike ZIF-8, which lost its long-range ordered structure upon high loading of I2, Cu-BTC retained its crystallinity at all I2 loadings, as indicated by synchrotron-based XRD and PDF analysis. TGA-MS analysis of I2 loaded Cu-BTC revealed no iodine or iodide species were released up to 150 °C, beyond which the evolution of iodide species was observed. I2 adsorption sites in Cu-BTC were explored by a combination of MD (molecular dynamics) simulations and Rietveld analysis, which revealed two preferred adsorption positions (Fig. 19b). One is in the smallest cage, and the other is located at the main pore. These adsorbed I2 molecules interact with the axial water molecules coordinated to the paddle wheel metal centers via tritopic van der Waals interactions or with the benzene tricarboxylate organic linker. It is interesting to see that I2 was preferentially adsorbed by hydrophilic Cu-BTC in the presence of moisture. The authors claim this is because the adsorbed I2 molecules form a hydrophobic barrier preventing additional adsorption of water molecules.

While some MOF materials show exceptionally high iodine capture capacity, and molecular iodine can be temporarily trapped in their pores through strong I2-MOF interaction or by aforementioned pressure-induced amorphization, further conversion of I2 loaded MOFs into long-term waste forms is imperative for safe sequestration of highly mobile iodine. In this context, Sava and co-workers140 explored the use of glass-composite material (GCM) as long-term waste forms for I2 loaded MOFs. Two different MOFs were selected for the study, ZIF-8 and HKUST-1, which were chosen for their low cost, high stability, and remarkable iodine capture capability. The authors first loaded iodine into the two MOFs through vapor adsorption at 75 °C (I2 vapor pressure of 0.014 atm). 120 and 150 wt% of iodine was loaded for ZIF-8 and HKUST-1, respectively. The I2 loaded MOFs were subsequently incorporated into low-sintering glass powder with the addition of silver flakes as an additional iodine scavenger. Glass-composite material (GCM) was formed at a sintering temperature of 500–525 °C. A typical GCM waste form was comprised of 80 wt% glass powder, 10 wt% I2 loaded MOF, and 10 wt% silver flakes. The resulting GCMs exhibited exceptional thermal ad chemical stability, and no release of iodine was detected during the sintering process or in the subsequent leach durability tests. Thus the incorporation of I2 loaded MOFs into monolithic, highly robust GCM waste forms is an effective way for long-term sequestration of volatile iodine.

Further improved I2 capture capacity has been achieved by tailor-made materials. In a recent study, Zhang and co-workers146 reported a record high I2 uptake of 216 wt% by a NbO type MOF, Zn2(tptc)(apy)(H2O) (tptc4− = terphenyl-3,3′′,5,5′′-tetracarboxylate, apy = aminopyridine). It has a BET surface area of 1470 m2 g−1, with aminopyridyl groups decorating the pore space. The authors attributed the exceptionally high I2 adsorption capacity of this compound to its large porosity, plentiful phenyl rings, and strong electron-donor amino groups. Other than MOFs, some pure-organic based porous materials have also been designed for I2 capture. Zhu and co-workers155 designed and synthesized a series of charged porous aromatic frameworks (PAFs) by Sonogashira–Hagihara coupling reactions of lithium tetrakis(4-iodophenyl)borate and different alkyne monomers. These PAFs showed super high iodine adsorption capacity (>250 wt% at 75 °C and ambient pressure), owing to multiple interactions between iodine and the charged aromatic frameworks with conjugated π-electrons. X-ray photoelectron spectroscopy (XPS) analysis revealed the encapsulated I2 in the pores exist as ionic I3, indicating the involvement of chemical reactions during iodine adsorption. Iodine adsorption performance in MOFs and other selected sorbents has been summarized in Table 2.

Table 2 Iodine adsorption by selected MOFs and porous organic materials
Adsorbent BET surface area (m2 g−1) Loading method Temperature (°C) I2 capacity (wt%) Ref.
UiO-66-PYDC 1030 I2/cyclohexane solution 25 125 141
Zn3(DL-lac)2(pybz)2 763 I2/hexane solution 25 101 134
TMU-16 I2/hexane solution 25 45 142
Zn9(btc)4(atz)12 229 I2/cyclohexane solution 25 40 143
Cu2I2(tppe)-1 303 I2/cyclohexane solution 25 32 101
Cu2I2(tppe)-3 285 I2/cyclohexane solution 25 30 101
JLU-Liu32 I2/cyclohexane solution 25 29 144
Cd(2-NH2bdc)(4-bpmh) 30 I2/hexane solution 25 28 145
Cu2I2(tppe)-2 320 I2/cyclohexane solution 25 27 101
JLU-Liu31 1700 I2/cyclohexane solution 25 25 144
Cd(bdc)(4-bpmh) 36 I2/hexane solution 25 14 145
Zn2(tptc)(apy)(H2O) 168 Iodine vapor 75 216 146
HKUST-1 1798 Iodine vapor (3.5 RH%) 75 175 138
ZIF-8 1810 Iodine vapor 75 125 135
Ni4(44pba)8 Iodine vapor 25 110 147
MOC-19 Iodine vapor 25 50 148
Cu4Cl3(TPVS)4(H2O)2 Iodine vapor 70 49 149
(CuI)2(tppe) Iodine vapor 20 45 150
Fe3(HCOO)6 385 Iodine vapor 25 35 151
Cd(L)(ClO4)2 Iodine vapor 25 32 152
Zn3(BTC)2(TIB)2 Iodine vapor 70 14 153
Cu4I4-MOF 641 Iodine vapor 25 13 154
Cd3(BTC)2(TIB)2 Iodine vapor 70 3 153
PAF-24 136 Iodine vapor 75 276 155
PAF-23 82 Iodine vapor 75 271 155
PAF-25 262 Iodine vapor 75 260 155
Azo-Trip 510 Iodine vapor 77 238 156
SCMP-2 855 Iodine vapor 77 222 157
CMP-E1 1213 Iodine vapor 75 215 158
CMPN-3 1368 Iodine vapor 70 208 159
NiP-CMP 2600 Iodine vapor 77 202 160
SCMP-1 413 Iodine vapor 77 188 157
PAF-1 5600 Fixed pressure (40 Pa) 25 186 161
PAF-21 Iodine vapor 75 152 155
JUC-Z2 2081 Fixed pressure (40 Pa) 25 144 161
CMPN-2 339 Iodine vapor 70 110 159
CMPN-1 230 Iodine vapor 70 97 159

3.2.2. Capture of organic iodides. Radioactive organic iodides (ROIs) including CH3I, CH3CH2I etc. are another type of species commonly present in the off-gas streams of nuclear waste. Similar to the capture of iodine, current technology for the removal of radioactive organic iodides also involves the use of silver impregnated/exchanged solid sorbents such as silica, alumina, and zeolites.162–164 As mentioned earlier, this type of material suffers from high cost associated with the use of a noble metal and poor recyclability. Triethylenediamine (TED) impregnated activated carbon (AC) has also been proposed for this application.162 However, the capture of radioactive organic iodides is often performed at high temperature (e.g. 150 °C) in order to facilitate the desired chemical reactions and eliminate/diminish the impact from water adsorption. As AC-based materials are usable only under 120 °C, they are not suitable for these conditions. In addition, the presence of NOx in the off-gas streams complicates the use of AC-based sorbents because of its low ignition temperature and the explosives that may form. Despite the fact that MOFs have been extensively explored for iodine capture as outlined above, their use for effectively trapping ROIs has not been exploited until the very recent work by our group. Inspired by the use of TED impregnated AC for the capture of ROIs, we designed a MOF-based molecular trap which exhibited exceptional ROI capture performance under both dry and simulated off-gas conditions (Fig. 20).67 The molecular trap was achieved by incorporating tertiary amines (TED and hexamethylenetetramine, HMTA) into a highly porous and robust MOF, MIL-101-Cr, via a coordination bond between N and the open Cr centers. The design rational is that the tertiary amine is grafted to the MOF framework by a single nitrogen atom, with the other end of the molecule (a N atom as well) decorating the pore surface and available as binding sites for organic iodides. The amine-functionalized materials, MIL-101-Cr-TED and MIL-101-Cr-HMTA, retained the same framework integrity as the original MOF and remained highly stable. The porosity of these two compounds showed a moderate decrease compared to the pristine structure, with BET surface areas of 2282 and 2272 m2 g−1 (vs. 3342 m2 g−1 for MIL-101-Cr) for the TED and HMTA analogues, respectively. Despite this minor decrease, both are still much more porous than any other benchmark adsorbents which have surface areas of 300–1000 m2 g−1. The ROI uptake capacity for these tailor-made adsorbents was evaluated by vapor adsorption. At 30 °C, MIL-101-Cr-TED and MIL-101-Cr-HMTA adsorb 166 and 174 wt% of CH3I respectively under a partial pressure of 0.2 atm. The adsorbed amounts are significantly higher than any other materials, which typically have uptakes of <60 wt% under the same conditions. CH3I adsorption experiments were also performed at 150 °C, which is more relevant to the capture of organic iodides from off-gas mixtures, and the CH3I uptake amounts were 71 and 62 wt% for TED and HMTA impregnated MOFs, respectively. This also greatly outperformed other sorbent materials such as TED@AC, HMTA@AC, Ag+@ZSM-5, and Ag+@MOR. The exceptionally high CH3I adsorption capacity of the tertiary amine functionalized MIL-101-Cr analogues was attributed to their high porosity and the effective grafting of TED and HMTA onto the pore surface, creating a molecular trap that greatly enhances the interaction with CH3I. The adsorption performance for other organic iodides including CH3CH2I and CH3CH2CH2I was also evaluated, and similar uptake capacities to CH3I were observed. The authors further explored the material's capability of capturing ROIs under real-world conditions by performing column breakthrough experiments of a simulated off-gas mixture including CH3I, I2, HNO3, and NOx under high humidity (95% RH) at 150 °C. The total iodine uptake amounts are 38 and 33 wt% for MIL-101-Cr-TED and MIL-101-Cr-HMTA, respectively, which are more than twice higher than that of HISL (16 wt%) and Ag0@MOR (5 wt%) under the same conditions. These results demonstrate that the amine grafted MIL-101-Cr materials hold great promise for effective capture of iodine and organic iodides from off-gas streams. The ROI capture mechanism was investigated by various techniques including HRTEM-EDS, solid-state 1H NMR, XPS, and in situ FT-IR, as well as theoretical calculations which support the proposed mechanism that the tertiary amines form strong chemical bonds with RI (R = –CH3, –CH2CH3, or –CH2CH2CH3) yielding ionic species (TED/HMTA-R)+ I at high temperatures. In a follow-up work,165 we carried out a systematic study to investigate the effect of different amines on ROI capture. The selected amines included TED, HMTA, N,N-dimethylethylenediamine (DMEDA), N,N-dimethyl-1,3-propanediamine (DMPDA), and N,N-dimethyl-1,4-butanediamine (DMBDA). Adsorption results show that MIL-101-Cr-DMEDA gave the highest uptake amount (80 wt%) among the five amine functionalized compounds, which also represents a record-high value for all reported sorbents. This can be attributed to its relatively high surface area (2460 m2 g−1) resulting from the smaller size of the functional amine (DMEDA). Remarkably, these amine functionalized MOF molecular traps can be recycled without loss of adsorption capacity.
image file: c7cs00885f-f20.tif
Fig. 20 (a) CH3I adsorption curves in MIL-101-Cr-TED, MIL-101-Cr-HMTA, and selected benchmark sorbents at 150 °C with partial pressure of 0.2 atm. (b) The recyclability of MIL-101-Cr-TED for CH3I adsorption. Adsorption: 150 °C, 0.2 atm of CH3I in nitrogen, desorption: removal of CH3I together with TED and re-grafting TED; (c) the CH3I uptake at 150 °C under dry and humid (81% RH) conditions by breakthrough experiment. (Back row: dry conditions; front row: humid conditions), (insert) the uptake drop ratio by comparing the CH3I uptake of dry and humid conditions. (d) Decontamination factors of CH3I by MIL-101-Cr-TED under simulated conditions representing gas mixtures produced during CH3I reprocessing, which include CH3I (50 ppm), H2O, HNO3, NO2, and NO at 150 °C.67 Reproduced from ref. 67 with permission from the Nature Publishing Group, copyright 2017.
3.2.3. Capture of noble gases. Other than radioactive iodine and organic iodides, the off-gas streams also contain noble gas radionuclides (predominately isotopes of Xe and Kr) which must be captured and sequestrated.16685Kr is the larger concern because of its long half-life (t1/2 = 10.8 years).

The largest challenge to the adsorptive capture of noble gases is their inert nature. Traditional sorbents such as zeolites and activated carbons have been tested for noble gases capture and sequestration,37 but they are generally plagued by low adsorption capacity owing to the inability to tune their pore geometry and functionality. MOFs hold great promise for the adsorption of noble gases because of their highly tunable structures and functionalities. To date, a variety of MOF materials have been studied for noble gases adsorption, and the progress in the field was recently reviewed.167 Nevertheless, research has so far mostly been focused on the separation of Xe/Kr binary mixtures at relatively high partial pressure (>0.1 bar), which is associated with Xe purification during air separation.166,168,169 In contrast, the capture of Xe or Kr from air (related to nuclear waste sequestration) has been rarely reported. In an early report, Liu and co-workers170 investigated the capture capability of noble gases on HKUST-1 and NiDOBDC and found that NiDOBDC can adsorb and separate Xe/Kr from a simulated off-gas stream containing 400 ppm Xe, 40 ppm Kr, 78% N2, 21% O2, 0.9% Ar, and 0.03% CO2. In a more recent study, Banerjee and co-workers166 studied noble gases adsorption on a calcium based MOF, Ca(SDB) (SDB = sulfonyldibenzoate), which was identified as the most selective material for Xe/Kr adsorption out of a set of 125[thin space (1/6-em)]000 MOF structures through the use of a high-throughput computational screening method that modeled their performance. Experimental results confirm that Ca(SDB) shows the largest Henry coefficient of Xe adsorption (38 mmol g−1 bar−1) and the highest Xe/Kr selectivity (derived from Henry coefficients) among all reported materials. The high adsorption affinity toward noble gases on Ca(SDB) was attributed to its tailored pore size. Its Xe and Kr capture capability under dilute conditions was evaluated by column breakthrough experiments with a representative gas mixture (400 ppm Xe, 40 ppm Kr, 78.1% N2, 20.9% O2, 0.03% CO2 and 0.9% Ar). The breakthrough curve shows that Kr and Xe are retained in the column for a longer time compared to other gas components. This is especially true of Xe, which exhibits a breakthrough time of more than 1 hour (Fig. 21). The breakthrough experiment was also conducted under 42% relative humidity, and the adsorption capacity of Xe and Kr was mostly retained in the presence of water vapor.

image file: c7cs00885f-f21.tif
Fig. 21 Column breakthrough experiments using CaSDB at room temperature and 1 atm. Inlet is a gas mixture with 400 ppm. Xe and 40 ppm. Kr balanced with air under (a) dry and (b) 42% relative humidity.166 Reproduced from ref. 166 with permission from the Nature Publishing Group, copyright 2016.

3.3. Capture of toxic industrial gases and chemical warfare agents

Toxic industrial gases (TIGs, Table 3) such as COx, NH3, H2S, SOx, and NOx are of particular concern because they are commonly involved in various industrial processes and are ubiquitous in the atmosphere. Removal of TIGs involves adsorption of target gases under varied situations. For example, as these toxic gases are used or produced in many applications, industrial and emergency workers are at risk of exposure in the event of any accidental spills or leaks. Thus, efficient adsorbent materials for their capture are needed for protection or precaution. Additionally, certain toxic species are contained in flue gas or engine exhaust and must be removed before being released into the air. On the other hand, the direct capture of toxic gases from air is needed in areas where their concentration is above exposure limits. Compared to TIGs, CWAs are less common as a threat, since they are not as readily available as the former. However, due to their exceptionally high toxicity, capture and detoxification of CWAs is of utmost importance. CWAs are typically less volatile compared to other toxic and hazardous species; however, their vapors are still a serious threat as they are highly toxic. The removal of TIGs and CWAs by MOFs has become a burgeoning research area over the past few years, and related advancements have been reviewed recently from various perspectives.32,33,71,172,173 We will therefore focus on giving a general summary of material design strategies and highlighting significant progresses reported recently.
Table 3 Physical properties and toxicity data for selected industrial hazardous gases. 8 hour PEL represents 8 hour time-weighted average (TWA) permissible exposure limit set by Occupational Safety and Health Administration (OSHA)
Gas Boiling point (K) Kinetic diameter (Å) 8 hour PEL (ppm)
CO2 216.6 3.3 5000
CO 81.6 3.7 50
NH3 239.8 2.9 50
H2S 212.8 3.6 20
NO2 294.3 4–5 5
NO 121 3.5 25
SO2 263 4.1 5

3.3.1. Capture of COx. Carbon dioxide (CO2) has a IDLH (Immediately Dangerous to Life and Health) value of 40[thin space (1/6-em)]000 ppm, which is much higher than other hazardous gases such as NH3 (300 ppm) and H2S (100 ppm) and indicates its low toxicity. However, the environmental threat of CO2 is not less than any other chemical, due to its implication in global warming as the primary greenhouse gas. Since CO2 is predominantly emitted from the combustion of fossil fuels, implementation of CO2 capture technologies to power plant flue gas could effectively lower the rising level of atmospheric CO2. Adsorptive capture of CO2 from flue gas by MOFs has been proposed as an alternative to the current capture technology involving the use of aqueous alkanolamine solutions, with the goal of at lowering the associated energy penalty. A more challenging application is direct air capture from the atmosphere. Although this is of prime importance, it is rarely addressed because of the low CO2 partial pressures involved (<420 ppm).174,175 Tremendous effort has been made to develop MOFs for capturing CO2 with high performance, and CO2 adsorption is probably the most extensively explored application for MOFs. As a result, significant progress has been constantly reported, and MOF-based CO2 capture technology holds promise for implementation in real-world systems. MOFs possessing open metal sites (OMSs) or with polar functional groups such as amines have shown preeminent CO2 capture performance with respect to adsorption selectivity and capacity.34 A prototype example is MOF-74 and its analogous/derivative materials. Long et al.81,176 developed mmen-Mg2(dobpdc) by amine grafting on the coordinatively unsaturated Mg2+ sites of Mg2(dobpdc), an expanded MOF-74-Mg analogue. This material is able to adsorb 3.14 mmol g−1 (12.1 wt%) of CO2 at 0.15 bar and 40 °C, conditions associated with CO2 capture from flue gas. Importantly, it is stable in the presence of water vapor and retains its affinity for CO2 under humid conditions. Additionally, mmen-Mg2(dobpdc) adsorbs 2.0 mmol g−1 (8.1 wt%) of CO2 at 25 °C at partial pressure as low as 0.39 mbar, suggesting its applicability for direct CO2 capture from air. A mechanistic study revealed that the adsorption proceeds by cooperative insertion of CO2 molecules into metal–amine bonds, leading to the formation of ammonium carbamate. Despite the chemisorption involved, the adsorbent is recyclable through temperature swings, with a regeneration energy appreciably lower than that of the aqueous amine solutions. In another work, through a crystal engineering or reticular chemistry strategy, Zaworotko, Eddaoudi and co-workers171,177 developed a series of tailored microporous MOFs built on coordinatively saturated metal centers, periodically arrayed hexafluorosilicate (SiF62−) anions, and pyridyl ligands. In spite of the absence of OMSs or amine groups, these materials exhibit exceptional CO2 capture performance even at very low pressure (Fig. 22), attributed to their controllable pore size and favorable electrostatic interactions afforded by the SiF62− anions. The heats of adsorption for CO2 on these materials are lower than that of MOFs with OMSs, indicating a lower energy penalty associated with adsorbent regeneration. Remarkably, their CO2 adsorption selectivity and capability is negligibly influenced by the presence of moisture. More recently, CO2 capture by MOF-based membranes has attracted considerable research interest due to its high energy efficiency, low maintenance, and ease of processing. Significant advances have been made for developing both pure MOF membranes and mixed matrix membranes, as recently reviewed by Balbuena and co-workers.178,179
image file: c7cs00885f-f22.tif
Fig. 22 (a) Crystal structure of SIFSIX-3-Cu. Color code: pyrazine: blue polygon, Cu: purple polyhedral, Si: light blue spheres, F: light green spheres. (b) CO2 adsorption isotherms at variable temperatures for SIFSIX-3-Cu. (c) CO2 volumetric uptake for SIFSIX-3-Cu at 298 K and low pressure, compared with SIFSIX-3-Zn, SIFSIX-2-Cu-I, and Mg-MOF-74. (d) Column breakthrough test for CO2/N2: 1000 ppm/99.9% for SIFSIX-3-Cu in dry as well as at 74% RH.171 Reproduced from ref. 171 with permission from the Nature Publishing Group, copyright 2014.

Carbon monoxide (CO) predominantly arises from the incomplete combustion of carbon-containing fuels. CO has a IDLH of 1200 ppm and threatens human health by causing tissue hypoxia. CO exposure is responsible for approximately 500 deaths annually. Research on CO adsorption by MOFs are mostly focused on those with OMSs such as MOF-74-M180 (M = Zn2+, Ni2+, Co2+, Mg2+, Mn2+, Fe2+), HKUST-1,181 and MIL-101-Cr182 due to their favorable coordination to CO molecules. Long and co-workers84 reported CO adsorption in a Fe-triazolate-based MOF, Fe-BTTri, with coordinatively unsaturated Fe2+ centers. It displays exceptional CO uptake capacity at low pressure (1.45 mmol g−1 at 100 μbar and ambient temperature) and high adsorption selectivity over H2, N2, CO2, and various hydrocarbons (Fig. 23). The adsorption mechanism involves a spin state transition of the Fe2+ centers from high spin to low spin upon CO coordination. In a follow-up work,85 the authors report two similar compounds where the iron(II) centers are linked into chains rather than the discrete nodes of the previous structure. The new structures, Fe2Cl2(bbtc) and Fe2Cl2(btdd), adsorb CO through a cooperative spin transition mechanism. They exhibit a large CO working capacity with low regeneration energies and high selectivities over other gases which may enable the adsorbent for CO extraction from industrial waste feeds.

image file: c7cs00885f-f23.tif
Fig. 23 (a) Crystal structure of Fe-BTTri. (b) Carbon monoxide adsorption isotherms measured at various temperatures in Fe-BTTri. (c) Adsorption isotherms of various gases collected at 25 °C for Fe-BTTri. (d) IAST selectivities for CO/H2, CO/N2, and CO/CH4 at varying concentrations at 25 °C and 1 bar of total pressure in Fe-BTTri.84 Reproduced from ref. 84 with permission from the American Chemical Society, copyright 2016.
3.3.2. Capture of ammonia. Ammonia (NH3) has an IDLH value of 300 ppm; exposure to high concentrations of ammonia may cause lung damage. The 15 minute and 8 hour exposure limits for ammonia are 35 and 25 ppm, respectively, set by OSHA (US Occupational Safety and Health Administration). Ammonia is widely used as fertilizer, cleaner, and chemical feedstock, and it has a global production of 176 million tons in 2014. Ammonia spills are frequent, threatening industrial and emergency workers. MOFs have been investigated for ammonia capture because the currently employed carbon-based adsorbents generally suffer from low capacity. Various computational and experimental studies demonstrate that MOFs possessing OMSs or functional groups that form strong hydrogen bonds with ammonia display high adsorption capacity at low pressure. However, many of these materials undergo loss of crystallinity and porosity upon exposure to ammonia of relatively high concentration, due to exceptionally strong adsorbate–adsorbent interaction.99 Dinca and co-workers104 recently reported ammonia adsorption by a series of mesoporous MOFs, M2Cl2(BTDD)(H2O)2 (M = Mn, Co, and Ni), built on bisbenzene–triazolate ligands. Despite the presence of OMSs, these materials exhibit high and reversible ammonia uptake (Fig. 24). They are able to adsorb 12.02–15.47 mmol g−1 of ammonia at room temperature and 1 bar, and no loss of capacity was observed after three adsorption–desorption cycles. Although some MOFs have shown high ammonia adsorption capacities, competitive binding with water is a challenging issue which has not been well addressed. To this end, the use of hydrophobic MOFs may be a possible solution as demonstrated by computational studies. By GCMC simulations, Ghosh and co-workers105 show that the NH3 uptake capacities of hydrophobic MOFs do not suffer a dramatic drop in the presence of water. This is because ammonia molecules are in close contact with the pore surface while water molecules are in the middle of the pores due to the hydrophobicity of the walls.
image file: c7cs00885f-f24.tif
Fig. 24 (a) Crystal structure of M2Cl2(BTDD)(H2O)2 (M = Mn, Co, and Ni) (b) NH3 adsorption (solid symbols) and desorption (open symbols) for Mn (red squares), Co (blue triangles), and Ni (green pentagons) analogues, compared with UiO-66-NH2 (grey circles).104 Reproduced from ref. 104 with permission from the American Chemical Society, copyright 2016.
3.3.3. Capture of H2S. Hydrogen sulfide (H2S) is a highly toxic and corrosive gas with a characteristic smell of rotten eggs and an odor threshold of as low as 0.47 ppb. OSHA has set a 8 hour permissible exposure limit of 10 ppm for H2S, above which it may cause eye irritation and damage. H2S is mainly produced in oil refinement and natural gas production, and unacceptable amounts of H2S must be removed from related gas streams. Various techniques have been applied to H2S capture, including absorption in polar liquids and by porous solids such as carbons and zeolites.183 MOFs have also been extensively explored for H2S capture, and some of the MOFs studied have shown very high capacity. For example, MIL-101 can take up 38 mmol g−1 of H2S at 20 bar due to its large pore volume.184 However, in real-world systems, H2S capture involves the selective removal of the molecule from gas streams containing CO2, CH4 and other gases from ppm to percentage levels. Recently, Belmabkhout and co-workers185 reported a fine-tuned Ga-soc-MOF for H2S capture. The compound is built on Ga3O(COO)6 SBUs linked by abtc4− (3,3′,5,5′-zaobenzene tetracarboxylates) ligands. It has a BET surface area of 1350 m2 g−1 and shows a high H2S tolerance and stability. The authors evaluated its H2S removal capability by performing column breakthrough tests with a mixture of CO2/H2S/CH4: 5/5/90 at room temperature. The results display a substantially longer retention time for H2S (40 min g−1) than that of CH4 and CO2 (less than 5 min g−1), indicating the potential of this material for H2S removal from refinery-off gases and natural gases.
3.3.4. Capture of SO2 and NO2. SO2 and NO2 are highly toxic acid gases that may damage the respiratory system upon exposure. Moreover, when dissolved in atmospheric water, they can both form acid rain. SO2 and NO2 are mainly emitted from automobiles and coal-fired power plants. MOF-based adsorptive capture technique has been proposed as an alternative to absorption by basic solutions for the removal of acidic gases, as the latter produces a large amount of wastewater. MOFs with OMSs such as MOF-74 and HKUST-1 are shown to have relatively high uptake amount and adsorption affinities toward NO2 and SO2. Chabal and co-workers187 investigated the interaction of NO2 and SO2 with MOF-74-M (M = Zn, Mg, Ni, Co) by in situ infrared spectroscopy and ab initio DFT calculations. While both gases interact strongly with OMSs, they exhibit distinct adsorption mechanisms. The bonding of NO2 with metal centers leads to its dissociation, forming NO and NO3. As for SO2, though it shows a high binding energy of 90 kJ mol−1 with the OMSs, the adsorption does not involve its chemical dissociation. NO2 capture was also studied in stable Zr-based MOFs without OMSs, such as UiO-66 and UiO-67. Bandosz and co-workers188 reported the influence of pore size on NO2 adsorption. Under dry conditions, the smaller pore size of UiO-66 shows a positive impact on adsorption as a result of increased contact between the adsorbate and the pore walls. However, in the presence of moisture, the large pore size of UiO-67 enhances its adsorption of water, thus facilitating its uptake of NO2 owing to the formation of nitric and nitrous acids in the pores. More recently, Peterson and co-workers189 studied NO2 removal from air (with NO2 concentration of 500–700 ppm) by UiO-66-NH2 under both dry and humid conditions. The functional amine group is found to considerably enhance NO2 removal capability of the MOF. More importantly, NO2 adsorption on UiO-66-NH2 generates a substantially reduced amount of NO, compared to the activated carbon BPL which is the benchmark adsorbent for toxic gas filtration. Investigation of the adsorption mechanism revealed the formation of a diazonium ion on the aromatic ring of the MOF. Xing and co-workers186 reported the use of inorganic anion (SiF62−) pillared MOFs for SO2 capture. They show exceptional SO2 uptake capacity even at very low concentration (2.31 mmol g−1 at 0.002 bar, Fig. 25), which was attributed to the strong electrostatic interaction between the SO2 molecules and the anions/aromatic rings of the MOFs. Importantly, these materials exhibit highly favorable adsorption of SO2 over CO2, which is not achievable by MOFs with OMSs such as MOF-74, yet is very important for SO2 removal from flue gas due to the prevalence of CO2.
image file: c7cs00885f-f25.tif
Fig. 25 Experimental column breakthrough curves for (a) SO2/N2 (2000 ppm SO2) separations with SIFSIX-1-Cu, SIFSIX-2-Cu-i, and SIFSIX-3-Ni, and (b) SO2/CO2 (2000 ppm SO2) separations with SIFSIX-1-Cu and SIFSIX-2-Cu-i at 298 K and 1.01 bar. Cycling column breakthrough tests for CO2/SO2 (2000 ppm SO2) separations with (c) SIFSIX-2-Cu-i and (d) SIFSIX-1-Cu at 298 K and 1.01 bar (mixed gas flow: 14 mL min−1). In panel (a), open circles are for N2, and filled circles are for SO2. In panels (b–d), the open circles are for CO2, and the filled circles are for SO2. CA/C0, outletconcentration/feed concentration.186 Reproduced from ref. 186 with permission from the John Wiley & Sons, Inc., copyright 2017.
3.3.5. Capture of CWAs. Due to the extreme toxicity of chemical warfare agents (CWAs), their removal by MOFs have been typically done with simulant molecules. MOFs have been studied for not only the adsorptive capture but also the catalytic detoxification/degradation of CWAs, but this has been mostly performed in solution, which is out of the scope of this review.190 Recently, Frenkel and co-workers61 reported the capture and decomposition of a nerve-agent simulant, dimethyl methyl phosphonate (DMMP) by a series of Zr-based MOFs including UiO-66, UiO-67, MOF-808, and NU-1000. The authors show that these MOFs are able to adsorb DMMP from air. EXAFS analysis indicates DMMP interacts directly with the Zr centers in the framework, leading to the hydroxylation of MOFs and the decomposition of DMMP. Thus, the integrity of the MOF structures is affected upon adsorption/desorption of DMMP to different degrees.

4. Conclusions and outlook

Over the past decade, researchers have shown that the physical and chemical properties of MOFs can be modified by various strategies, including substituting metal nodes/organic likers, changing their connectivity, and post-synthetic functionalization, to name a few. The tremendous diversity and high tunability of MOFs allow the design of tailor-made materials with desirable features for specific applications. To this end, a considerable number of MOFs have been evaluated for their performance on sensing and/or capture of toxic and hazardous species. Some MOFs exhibit advantages over conventional materials in one or more aspects, however, some issues remain to be explored and addressed prior to their implementation in real-world applications.

Luminescence-based sensing has been a key area for LMOF research, and significant progress has been made in detecting nearly every class of compound.19 However, the great bulk of LMOF sensing has focused on analytes present in solution, rather than in the vapor or gas phase, despite the fact that many applications require detection to be done in the gas/vapor phase. Currently, the field of vapor and gas-phase sensing is under investigated. Even the subfields of vapor-phase sensing that have been explored, such as VOC sensing, need more careful analysis. Often the reported work give the percentage by which emission can be quenched or enhanced, but fail to specify the analyte concentration associated with the emission change, or to investigate the detection limit. While MOF stability has improved from its early stage through the use of high valence metals or hydrophobic ligands, it is not uncommon for reports of new LMOF sensors to omit information on the stability of the sensor under real-world conditions. Furthermore, PXRD measurements are the most commonly cited evidence for the stability of a MOF under a given set of conditions. However, this only gives a broad picture of the sample's large-scale crystallinity, and fails to address issues such as altered pore or grain boundary conditions which can greatly affect sensing and gas adsorption applications. Surface area measurements, rather than PXRD, could provide a more nuanced view of MOF stability for applications that rely on mass transfer through that MOF, and its use as a stability assay should be more widely adopted. Additionally, improving sensing selectivity and decreasing response time remain challenging tasks for future development. Gas sensing by selective absorbance and pre-concentration of an analyte resulting in a luminescence change would be a powerful technique, but reports of such materials are very limited.53 In many cases, this may be due to practical difficulties in combining gas adsorption and photoluminescence measurements.

With respect to the capture of hazardous gases and vapors, MOFs exhibit enormous potential for some specific applications, such as the capture of CO2 from flue gas and radioactive molecular iodide and organic iodides from nuclear waste. However, for numerous other areas, current research remains at early stages, and therefore, considerable efforts are needed to address the existing challenges. First of all, while some MOFs show very high uptake capacity and/or adsorption selectivity toward certain molecule, the carefully designed linkers required complicated organic synthesis. The use of these exceptionally expensive ligands impedes their further consideration and evaluation for real applications. In addition, despite the fact that the stability of MOFs has improved markedly over the past several years, and some MOFs such as MIL-101-Cr and UiO-66 have shown exceptional thermal and chemical robustness that is comparable to inorganic adsorbents, a significant number of MOFs still suffer from relatively poor stability, particularly in regards to the loss of crystallinity and porosity upon prolonged exposure to water/acidic vapors at elevated temperature. Moreover, many of the adsorption experiments have been tested under conditions that are not relevant to industrial applications with respect to temperature, pressure, impurities, etc. While these measurements can provide some useful information regarding the adsorption capacity/selectivity of the adsorbent, further investigations under conditions comparable to the real world applications are necessary to fully evaluate the usefulness of the material.

Despite these challenges, research on metal–organic frameworks has been one of the fastest growing fields in material chemistry and will continue to advance in the future. It offers incredible tunability for tailor-made material design and optimization, and true potential for applications in gas capture and sensing, as well as in many other areas.


abtc3,3′,5,5′-Zaobenzene tetracarboxylates
ACActivated carbon
Azo-TripAzolinked microporous polymer
4-bpmh N,N-Bis-pyridin-4-ylmethylene-hydrazine
BPP3,3′-Dioxido-4,4′-biphenyldicarboxylate, biphenyl with para-COOH
CMPConjugated microporous polymer
CMPNConjugated microporous polymer nanotubes
CWAChemical warfare agent
DL-lac DL-Lactic acid
DMBDA N,N-Dimethyl-1,4-butanediamine
DMEDA N,N-Dimethylethylenediamine
DMPDA N,N-Dimethyl-1,3-propanediamine
DMMPDimethyl methyl phosphonate
FRETFluorescence resonance energy transfer
GCMGlass-composite material
GCMCGrand canonical Monte Carlo
GWPGlobal warming potential
HISLHydrophobicity-intensified silicalite-1
IASTIdeal adsorbed solution theory
HOMOHighest occupied molecular orbital
L12,2′-Bis (trifluoromethyl)-4,4′-biphenyldicarboxylate
ldcLinear dicarboxylates
LLCTLigand-to-ligand charge transfer
LMCTLigand-to-metal charge transfer
LMOFLuminescent metal–organic framework
LUMOLowest unoccupied molecular orbital
MLCTMetal-to-ligand charge transfer
MOFMetal–organic framework
NACNitroaromatic compounds
oba4,4′-Oxybis(benzoic acid)
OMSOpen metal site
PAFPorous aromatic framework
PYDCPyridine-dicarboxylic acid
RHRelative humidity
ROIRadioactive organic iodide
SBUSecondary building unit
SCMPConjugated microporous polymers having thiophene building blocks
TPP3,3′-Dioxido-4,4′-triphenyldicarboxylate, triphenyl with para-COOH
tppe1,1,2,2-Tetrakis(4-(pyridin-4-yl)phenyl) ethane
VOCVolatile organic compound

Conflicts of interest

There are no conflicts to declare.


We would like to thank the US Department of Energy, Basic Energy Sciences, division of Materials Sciences and Engineering for support through Grant No. DE-FG02-08ER46491.

Notes and references

  1. S. R. Batten, N. R. Champness, X.-M. Chen, J. Garcia-Martinez, S. Kitagawa, L. Öhrström, M. O’Keeffe, M. Paik Suh and J. Reedijk, Pure Appl. Chem., 2013, 85, 1715–1724 CrossRef CAS .
  2. O. K. Farha, I. Eryazici, N. C. Jeong, B. G. Hauser, C. E. Wilmer, A. A. Sarjeant, R. Q. Snurr, S. T. Nguyen, A. Ö. Yazaydın and J. T. Hupp, J. Am. Chem. Soc., 2012, 134, 15016–15021 CrossRef CAS PubMed .
  3. H. Furukawa, K. E. Cordova, M. O’Keeffe and O. M. Yaghi, Science, 2013, 341 Search PubMed .
  4. Z.-F. Wu, B. Tan, M.-L. Feng, A.-J. Lan and X.-Y. Huang, J. Mater. Chem. A, 2014, 2, 6426 CAS .
  5. P. Silva, S. M. F. Vilela, J. P. C. Tome and F. A. Almeida Paz, Chem. Soc. Rev., 2015, 44, 6774–6803 RSC .
  6. J. A. Mason, J. Oktawiec, M. K. Taylor, M. R. Hudson, J. Rodriguez, J. E. Bachman, M. I. Gonzalez, A. Cervellino, A. Guagliardi, C. M. Brown, P. L. Llewellyn, N. Masciocchi and J. R. Long, Nature, 2015, 527, 357 CrossRef CAS PubMed .
  7. B. Li, H.-M. Wen, W. Zhou, J. Q. Xu and B. Chen, Chem, 2016, 1, 557–580 CAS .
  8. J. Sculley, D. Yuan and H.-C. Zhou, Energy Environ. Sci., 2011, 4, 2721–2735 CAS .
  9. M. P. Suh, H. J. Park, T. K. Prasad and D.-W. Lim, Chem. Rev., 2012, 112, 782–835 CrossRef CAS PubMed .
  10. J.-R. Li, J. Sculley and H.-C. Zhou, Chem. Rev., 2012, 112, 869–932 CrossRef CAS PubMed .
  11. Z. Bao, G. Chang, H. Xing, R. Krishna, Q. Ren and B. Chen, Energy Environ. Sci., 2016, 9, 3612–3641 CAS .
  12. H. Wu, Q. Gong, D. H. Olson and J. Li, Chem. Rev., 2012, 112, 836–868 CrossRef CAS PubMed .
  13. Z. R. Herm, E. D. Bloch and J. R. Long, Chem. Mater., 2014, 26, 323–338 CrossRef CAS .
  14. X. Cui, K. Chen, H. Xing, Q. Yang, R. Krishna, Z. Bao, H. Wu, W. Zhou, X. Dong, Y. Han, B. Li, Q. Ren, M. J. Zaworotko and B. Chen, Science, 2016, 353, 141–144 CrossRef CAS PubMed .
  15. L. Zhu, X.-Q. Liu, H.-L. Jiang and L.-B. Sun, Chem. Rev., 2017, 117, 8129–8176 CrossRef CAS PubMed .
  16. L. Zeng, X. Guo, C. He and C. Duan, ACS Catal., 2016, 6, 7935–7947 CrossRef CAS .
  17. Y.-B. Huang, J. Liang, X.-S. Wang and R. Cao, Chem. Soc. Rev., 2017, 46, 126–157 RSC .
  18. L. Ma, J. M. Falkowski, C. Abney and W. Lin, Nat. Chem., 2010, 2, 838 CrossRef CAS PubMed .
  19. W. P. Lustig, S. Mukherjee, N. D. Rudd, A. V. Desai, J. Li and S. K. Ghosh, Chem. Soc. Rev., 2017, 46, 3242–3285 RSC .
  20. Z. Hu, B. J. Deibert and J. Li, Chem. Soc. Rev., 2014, 43, 5815–5840 RSC .
  21. Z. Hu, W. P. Lustig, J. Zhang, C. Zheng, H. Wang, S. J. Teat, Q. Gong, N. D. Rudd and J. Li, J. Am. Chem. Soc., 2015, 137, 16209–16215 CrossRef CAS PubMed .
  22. M. D. Allendorf, C. A. Bauer, R. K. Bhakta and R. J. T. Houk, Chem. Soc. Rev., 2009, 38, 1330–1352 RSC .
  23. A.-L. Li, Q. Gao, J. Xu and X.-H. Bu, Coord. Chem. Rev., 2017, 344, 54–82 CrossRef CAS .
  24. F.-M. Zhang, L.-Z. Dong, J.-S. Qin, W. Guan, J. Liu, S.-L. Li, M. Lu, Y.-Q. Lan, Z.-M. Su and H.-C. Zhou, J. Am. Chem. Soc., 2017, 139, 6183–6189 CrossRef CAS PubMed .
  25. G. K. H. Shimizu, J. M. Taylor and S. Kim, Science, 2013, 341, 354–355 CrossRef CAS PubMed .
  26. P. Ramaswamy, N. E. Wong and G. K. H. Shimizu, Chem. Soc. Rev., 2014, 43, 5913–5932 RSC .
  27. Y. Zhao, Chem. Mater., 2016, 28, 8079–8081 CrossRef CAS .
  28. C. Pettinari, F. Marchetti, N. Mosca, G. Tosi and A. Drozdov, Polym. Int., 2017, 66, 731–744 CrossRef CAS .
  29. A. Cadiau, K. Adil, P. M. Bhatt, Y. Belmabkhout and M. Eddaoudi, Science, 2016, 353, 137–140 CrossRef CAS PubMed .
  30. Nat. Chem., 2016, 8, 987 Search PubMed.
  31. M. J. Clinton, N. Engl. J. Med., 1948, 238, 51–54 CrossRef CAS .
  32. J. B. DeCoste and G. W. Peterson, Chem. Rev., 2014, 114, 5695–5727 CrossRef CAS PubMed .
  33. P. Kumar, K.-H. Kim, E. E. Kwon and J. E. Szulejko, J. Mater. Chem. A, 2016, 4, 345–361 CAS .
  34. K. Sumida, D. L. Rogow, J. A. Mason, T. M. McDonald, E. D. Bloch, Z. R. Herm, T.-H. Bae and J. R. Long, Chem. Rev., 2012, 112, 724–781 CrossRef CAS PubMed .
  35. R. Quadrelli and S. Peterson, Energy Policy, 2007, 35, 5938–5952 CrossRef .
  36. M. Lenzen, Energy Convers. Manage., 2008, 49, 2178–2199 CrossRef CAS .
  37. N. R. Soelberg, T. G. Garn, M. R. Greenhalgh, J. D. Law, R. Jubin, D. M. Strachan and P. K. Thallapally, Sci. Technol. Nucl. Install., 2013, 2013, 12 Search PubMed .
  38. L. Szinicz, Toxicology, 2005, 214, 167–181 CrossRef CAS PubMed .
  39. D. C. Wang, Y. H. Li, D. Li, Y. Z. Xia and J. P. Zhang, Renewable Sustainable Energy Rev., 2010, 14, 344–353 CrossRef CAS .
  40. S. Srivastava, B. K. Gupta and R. Gupta, Cryst. Growth Des., 2017, 17, 3907–3916 CAS .
  41. L. Han, J. Zhou, X. Li, C.-Y. Sun, L. Zhao, Y.-T. Zhang, M. Zhu, X.-L. Wang and Z.-M. Su, Inorg. Chem. Commun., 2017, 86, 200–203 CrossRef CAS .
  42. X.-Y. Xu and B. Yan, Dalton Trans., 2016, 45, 7078–7084 RSC .
  43. Z. Dou, J. Yu, H. Xu, Y. Cui, Y. Yang and G. Qian, Microporous Mesoporous Mater., 2013, 179, 198–204 CrossRef CAS .
  44. J.-J. Liu, Y.-B. Shan, C.-R. Fan, M.-J. Lin, C.-C. Huang and W.-X. Dai, Inorg. Chem., 2016, 55, 3680–3684 CrossRef CAS PubMed .
  45. X.-Y. Xu and B. Yan, J. Mater. Chem. A, 2017, 5, 2215–2223 CAS .
  46. Z. Wei, Z.-Y. Gu, R. K. Arvapally, Y.-P. Chen, R. N. McDougald, J. F. Ivy, A. A. Yakovenko, D. Feng, M. A. Omary and H.-C. Zhou, J. Am. Chem. Soc., 2014, 136, 8269–8276 CrossRef CAS PubMed .
  47. Z. Hu, G. Huang, W. P. Lustig, F. Wang, H. Wang, S. J. Teat, D. Banerjee, D. Zhang and J. Li, Chem. Commun., 2015, 51, 3045–3048 RSC .
  48. J. Heine and K. Mueller-Buschbaum, Chem. Soc. Rev., 2013, 42, 9232–9242 RSC .
  49. B. Zhao, N. Li, X. Wang, Z. Chang and X.-H. Bu, ACS Appl. Mater. Interfaces, 2017, 9, 2662–2668 CAS .
  50. C. Zhang, L. Sun, Y. Yan, Y. Liu, Z. Liang, Y. Liu and J. Li, J. Mater. Chem. C, 2017, 5, 2084–2089 RSC .
  51. W. P. Lustig, F. Wang, S. J. Teat, Z. Hu, Q. Gong and J. Li, Inorg. Chem., 2016, 55, 7250–7256 CrossRef CAS PubMed .
  52. J. Zhang, W.-B. Yang, X.-Y. Wu, X.-F. Kuang and C.-Z. Lu, Dalton Trans., 2015, 44, 13586–13591 RSC .
  53. N. Yanai, K. Kitayama, Y. Hijikata, H. Sato, R. Matsuda, Y. Kubota, M. Takata, M. Mizuno, T. Uemura and S. Kitagawa, Nat. Mater., 2011, 10, 787–793 CrossRef CAS PubMed .
  54. R.-W. Huang, Y.-S. Wei, X.-Y. Dong, X.-H. Wu, C.-X. Du, S.-Q. Zang and T. C. W. Mak, Nat. Chem., 2017, 9, 689–697 CrossRef CAS PubMed .
  55. Y. Yu, J.-P. Ma, C.-W. Zhao, J. Yang, X.-M. Zhang, Q.-K. Liu and Y.-B. Dong, Inorg. Chem., 2015, 54, 11590–11592 CrossRef CAS PubMed .
  56. J. Li, X. Yu, M. Xu, W. Liu, E. Sandraz, H. Lan, J. Wang and S. M. Cohen, J. Am. Chem. Soc., 2017, 139, 611–614 CrossRef CAS PubMed .
  57. S. Chen, J. Zhang, T. Wu, P. Feng and X. Bu, J. Am. Chem. Soc., 2009, 131, 16027–16029 CrossRef CAS PubMed .
  58. J. Tang and J. Wang, RSC Adv., 2017, 7, 50829–50837 RSC .
  59. C. Yang, J. Cheng, Y. Chen and Y. Hu, Appl. Surf. Sci., 2017, 420, 252–259 CrossRef CAS .
  60. H. J. Park, J. K. Jang, S.-Y. Kim, J.-W. Ha, D. Moon, I.-N. Kang, Y.-S. Bae, S. Kim and D.-H. Hwang, Inorg. Chem., 2017, 56, 12098–12101 CrossRef CAS PubMed .
  61. A. M. Plonka, Q. Wang, W. O. Gordon, A. Balboa, D. Troya, W. Guo, C. H. Sharp, S. D. Senanayake, J. R. Morris, C. L. Hill and A. I. Frenkel, J. Am. Chem. Soc., 2017, 139, 599–602 CrossRef CAS PubMed .
  62. R. G. AbdulHalim, P. M. Bhatt, Y. Belmabkhout, A. Shkurenko, K. Adil, L. J. Barbour and M. Eddaoudi, J. Am. Chem. Soc., 2017, 139, 10715–10722 CrossRef CAS PubMed .
  63. Z. Li, A. W. Peters, A. E. Platero-Prats, J. Liu, C.-W. Kung, H. Noh, M. R. DeStefano, N. M. Schweitzer, K. W. Chapman, J. T. Hupp and O. K. Farha, J. Am. Chem. Soc., 2017, 139, 15251–15258 CrossRef CAS PubMed .
  64. A. H. Assen, Y. Belmabkhout, K. Adil, P. M. Bhatt, D.-X. Xue, H. Jiang and M. Eddaoudi, Angew. Chem., Int. Ed., 2015, 54, 14353–14358 CrossRef CAS PubMed .
  65. C.-X. Chen, Q.-F. Qiu, C.-C. Cao, M. Pan, H.-P. Wang, J.-J. Jiang, Z.-W. Wei, K. Zhu, G. Li and C.-Y. Su, Chem. Commun., 2017, 53, 11403–11406 RSC .
  66. Y. G. Chung, D. A. Gómez-Gualdrón, P. Li, K. T. Leperi, P. Deria, H. Zhang, N. A. Vermeulen, J. F. Stoddart, F. You, J. T. Hupp, O. K. Farha and R. Q. Snurr, Sci. Adv., 2016, 2, e1600909 Search PubMed .
  67. B. Li, X. Dong, H. Wang, D. Ma, K. Tan, S. Jensen, B. J. Deibert, J. Butler, J. Cure, Z. Shi, T. Thonhauser, Y. J. Chabal, Y. Han and J. Li, Nat. Commun., 2017, 8, 485 CrossRef PubMed .
  68. S. M. Cohen, Chem. Rev., 2012, 112, 970–1000 CrossRef CAS PubMed .
  69. D. Britt, D. Tranchemontagne and O. M. Yaghi, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 11623–11627 CrossRef CAS PubMed .
  70. I. Spanopoulos, P. Xydias, C. D. Malliakas and P. N. Trikalitis, Inorg. Chem., 2013, 52, 855–862 CrossRef CAS PubMed .
  71. N. S. Bobbitt, M. L. Mendonca, A. J. Howarth, T. Islamoglu, J. T. Hupp, O. K. Farha and R. Q. Snurr, Chem. Soc. Rev., 2017, 46, 3357–3385 RSC .
  72. A. C. Kizzie, A. G. Wong-Foy and A. J. Matzger, Langmuir, 2011, 27, 6368–6373 CrossRef CAS PubMed .
  73. C. Wang, X. Liu, N. Keser Demir, J. P. Chen and K. Li, Chem. Soc. Rev., 2016, 45, 5107–5134 RSC .
  74. M. Bosch, M. Zhang and H.-C. Zhou, Adv. Chem., 2014, 2014, 8 Search PubMed .
  75. Y. Bai, Y. Dou, L.-H. Xie, W. Rutledge, J.-R. Li and H.-C. Zhou, Chem. Soc. Rev., 2016, 45, 2327–2367 RSC .
  76. G. Férey, C. Mellot-Draznieks, C. Serre, F. Millange, J. Dutour, S. Surblé and I. Margiolaki, Science, 2005, 309, 2040–2042 CrossRef PubMed .
  77. J. H. Cavka, S. Jakobsen, U. Olsbye, N. Guillou, C. Lamberti, S. Bordiga and K. P. Lillerud, J. Am. Chem. Soc., 2008, 130, 13850–13851 CrossRef PubMed .
  78. Y. Huang, W. Qin, Z. Li and Y. Li, Dalton Trans., 2012, 41, 9283–9285 RSC .
  79. K. Wang, H. Huang, W. Xue, D. Liu, X. Zhao, Y. Xiao, Z. Li, Q. Yang, L. Wang and C. Zhong, CrystEngComm, 2015, 17, 3586–3590 RSC .
  80. Q. Sun, B. Aguila, G. Verma, X. Liu, Z. Dai, F. Deng, X. Meng, F.-S. Xiao and S. Ma, Chem, 2016, 1, 628–639 CAS .
  81. T. M. McDonald, J. A. Mason, X. Kong, E. D. Bloch, D. Gygi, A. Dani, V. Crocellà, F. Giordanino, S. O. Odoh, W. S. Drisdell, B. Vlaisavljevich, A. L. Dzubak, R. Poloni, S. K. Schnell, N. Planas, K. Lee, T. Pascal, L. F. Wan, D. Prendergast, J. B. Neaton, B. Smit, J. B. Kortright, L. Gagliardi, S. Bordiga, J. A. Reimer and J. R. Long, Nature, 2015, 519, 303 CrossRef CAS PubMed .
  82. P. J. Landrigan, R. Fuller, N. J. R. Acosta, O. Adeyi, R. Arnold, N. Basu, A. B. Baldé, R. Bertollini, S. Bose-O'Reilly, J. I. Boufford, P. N. Breysse, T. Chiles, C. Mahidol, A. M. Coll-Seck, M. L. Cropper, J. Fobil, V. Fuster, M. Greenstone, A. Haines, D. Hanrahan, D. Hunter, M. Khare, A. Krupnick, B. Lanphear, B. Lohani, K. Martin, K. V. Mathiasen, M. A. McTeer, C. J. L. Murray, J. D. Ndahimananjara, F. Perera, J. Potočnik, A. S. Preker, J. Ramesh, J. Rockström, C. Salinas, L. D. Samson, K. Sandilya, P. D. Sly, K. R. Smith, A. Steiner, R. B. Stewart, W. A. Suk, O. C. P. van Schayck, G. N. Yadama, K. Yumkella and M. Zhong, The Lancet, 2017 Search PubMed .
  83. 7 million premature deaths annually linked to air pollution, World Health Organization news release, 25 Mar 2014,, accessed 11/09, 2017.
  84. D. A. Reed, D. J. Xiao, M. I. Gonzalez, L. E. Darago, Z. R. Herm, F. Grandjean and J. R. Long, J. Am. Chem. Soc., 2016, 138, 5594–5602 CrossRef CAS PubMed .
  85. D. A. Reed, B. K. Keitz, J. Oktawiec, J. A. Mason, T. Runčevski, D. J. Xiao, L. E. Darago, V. Crocellà, S. Bordiga and J. R. Long, Nature, 2017, 550, 96 CAS .
  86. P. Kovacic and R. Somanathan, J. Appl. Toxicol., 2014, 34, 810–824 CrossRef CAS PubMed .
  87. J. M. D. S. Roy Harrison, F. Dor and R. Henderson, WHO Guidelines for Indoor Air Quality: Selected Pollutants, World Health Organization, Geneva, 2010 Search PubMed .
  88. A. Lan, K. Li, H. Wu, D. H. Olson, T. J. Emge, W. Ki, M. Hong and J. Li, Angew. Chem., Int. Ed., 2009, 48, 2334–2338 CrossRef CAS PubMed .
  89. A. Lan, K. Li, H. Wu, L. Kong, N. Nijem, D. H. Olson, T. J. Emge, Y. J. Chabal, D. C. Langreth, M. Hong and J. Li, Inorg. Chem., 2009, 48, 7165–7173 CrossRef CAS PubMed .
  90. Z. Hu, S. Pramanik, K. Tan, C. Zheng, W. Liu, X. Zhang, Y. J. Chabal and J. Li, Cryst. Growth Des., 2013, 13, 4204–4207 CAS .
  91. J. Qin, B. Ma, X.-F. Liu, H.-L. Lu, X.-Y. Dong, S.-Q. Zang and H. Hou, J. Mater. Chem. A, 2015, 3, 12690–12697 CAS .
  92. F. Wang, C. Dong, Z. Wang, Y. Cui, C. Wang, Y. Zhao and G. Li, Eur. J. Inorg. Chem., 2014, 6239–6245 CrossRef CAS .
  93. S. Pramanik, C. Zheng, X. Zhang, T. J. Emge and J. Li, J. Am. Chem. Soc., 2011, 133, 4153–4155 CrossRef CAS PubMed .
  94. D. Banerjee, Z. Hu and J. Li, Dalton Trans., 2014, 43, 10668–10685 RSC .
  95. D. Banerjee, Z. C. Hu, S. Pramanik, X. Zhang, H. Wang and J. Li, CrystEngComm, 2013, 15, 9745–9750 RSC .
  96. S. Pramanik, Z. C. Hu, X. Zhang, C. Zheng, S. Kelly and J. Li, Chem. – Eur. J., 2013, 19, 15964–15971 CrossRef CAS PubMed .
  97. F.-Y. Yi, Y. Wang, J.-P. Li, D. Wu, Y.-Q. Lan and Z.-M. Sun, Mater. Horiz., 2015, 2, 245–251 RSC .
  98. A. H. Goldstein and I. E. Galbally, Environ. Sci. Technol., 2007, 41, 1514–1521 CrossRef CAS PubMed .
  99. L. Huang, T. Bandosz, K. L. Joshi, A. C. T. V. Duin and K. E. Gubbins, J. Chem. Phys., 2013, 138, 034102 CrossRef PubMed .
  100. H. Zhang, J. Yang, Y.-Y. Liu, S. Song and J.-F. Ma, Cryst. Growth Des., 2016, 16, 3244–3255 Search PubMed .
  101. S.-S. Zhao, L. Chen, X. Zheng, L. Wang and Z. Xie, Chem. – Asian J., 2017, 12, 615–620 CrossRef CAS PubMed .
  102. P. Mani, A. A. Ojha, V. S. Reddy and S. Mandal, Inorg. Chem., 2017, 56, 6772–6775 CrossRef CAS PubMed .
  103. Y. Li, S. Zhang and D. Song, Angew. Chem., Int. Ed., 2013, 52, 710–713 CrossRef CAS PubMed .
  104. A. J. Rieth, Y. Tulchinsky and M. Dincă, J. Am. Chem. Soc., 2016, 138, 9401–9404 CrossRef CAS PubMed .
  105. P. Ghosh, K. C. Kim and R. Q. Snurr, J. Phys. Chem. C, 2014, 118, 1102–1110 CAS .
  106. Z. Bai, Y. Dong, Z. Wang and T. Zhu, Environ. Int., 2006, 32, 303–311 CrossRef CAS PubMed .
  107. G. Nickerl, I. Senkovska and S. Kaskel, Chem. Commun., 2015, 51, 2280–2282 RSC .
  108. J.-N. Hao and B. Yan, Nanoscale, 2016, 8, 2881–2886 RSC .
  109. N. B. Shustova, A. F. Cozzolino, S. Reineke, M. Baldo and M. Dinca, J. Am. Chem. Soc., 2013, 135, 13326–13329 CrossRef CAS PubMed .
  110. C.-W. Zhao, J.-P. Ma, Q.-K. Liu, X.-R. Wang, Y. Liu, J. Yang, J.-S. Yang and Y.-B. Dong, Chem. Commun., 2016, 52, 5238–5241 RSC .
  111. C. Montoro, F. Linares, E. Quartapelle Procopio, I. Senkovska, S. Kaskel, S. Galli, N. Masciocchi, E. Barea and J. A. R. Navarro, J. Am. Chem. Soc., 2011, 133, 11888–11891 CrossRef CAS PubMed .
  112. M. J. Katz, J. E. Mondloch, R. K. Totten, J. K. Park, S. B. T. Nguyen, O. K. Farha and J. T. Hupp, Angew. Chem., Int. Ed., 2014, 53, 497–501 CrossRef CAS PubMed .
  113. J. E. Mondloch, M. J. Katz, W. C. Isley, III, P. Ghosh, P. Liao, W. Bury, G. W. Wagner, M. G. Hall, J. B. De Coste, G. W. Peterson, R. Q. Snurr, C. J. Cramer, J. T. Hupp and O. K. Farha, Nat. Mater., 2015, 14, 512–516 CrossRef CAS PubMed .
  114. S.-Y. Moon, A. J. Howarth, T. Wang, N. A. Vermeulen, J. T. Hupp and O. K. Farha, Chem. Commun., 2016, 52, 3438–3441 RSC .
  115. S.-R. Zheng, R.-L. Chen, Z.-M. Liu, X.-L. Wen, T. Xie, J. Fan and W.-G. Zhang, CrystEngComm, 2014, 16, 2898–2909 RSC .
  116. B. R. Varju, J. S. Ovens and D. B. Leznoff, Chem. Commun., 2017, 53, 6500–6503 RSC .
  117. J. S. Ovens, P. R. Christensen and D. B. Leznoff, Chem. – Eur. J., 2016, 22, 8234–8239 CrossRef CAS PubMed .
  118. D. A. Giannakoudakis, Y. Hu, M. Florent and T. J. Bandosz, Nanoscale Horiz., 2017, 2, 356–364 RSC .
  119. W.-C. Li, H. Bai, J.-N. Hsu, S.-N. Li and C. Chen, Ind. Eng. Chem. Res., 2008, 47, 1501–1505 CrossRef CAS .
  120. M. Ozekmekci, G. Salkic and M. F. Fellah, Fuel Process. Technol., 2015, 139, 49–60 CrossRef CAS .
  121. J. Kazmierczak-Razna, P. Nowicki and R. Pietrzak, Chem. Eng. Res. Des., 2016, 109, 346–353 CrossRef CAS .
  122. C. C. Rodrigues, D. de Moraes, S. W. da Nóbrega and M. G. Barboza, Bioresour. Technol., 2007, 98, 886–891 CrossRef CAS PubMed .
  123. V. Ochoa-Herrera and R. Sierra-Alvarez, Chemosphere, 2008, 72, 1588–1593 CrossRef CAS PubMed .
  124. J. C. Laube, M. J. Newland, C. Hogan, C. A. M. Brenninkmeijer, P. J. Fraser, P. Martinerie, D. E. Oram, C. E. Reeves, T. Röckmann, J. Schwander, E. Witrant and W. T. Sturges, Nat. Geosci., 2014, 7, 266 CrossRef CAS .
  125. R. K. Motkuri, H. V. R. Annapureddy, M. Vijaykumar, H. T. Schaef, P. F. Martin, B. P. McGrail, L. X. Dang, R. Krishna and P. K. Thallapally, Nat. Commun., 2014, 5, 4368 CAS .
  126. T.-H. Chen, I. Popov, W. Kaveevivitchai, Y.-C. Chuang, Y.-S. Chen, O. Daugulis, A. J. Jacobson and O. Š. Miljanić, Nat. Commun., 2014, 5, 5131 CrossRef CAS PubMed .
  127. T.-H. Chen, I. Popov, W. Kaveevivitchai, Y.-C. Chuang, Y.-S. Chen, A. J. Jacobson and O. Š. Miljanić, Angew. Chem., Int. Ed., 2015, 54, 13902–13906 CrossRef CAS PubMed .
  128. R.-B. Lin, T.-Y. Li, H.-L. Zhou, C.-T. He, J.-P. Zhang and X.-M. Chen, Chem. Sci., 2015, 6, 2516–2521 RSC .
  129. C.-X. Chen, Z. Wei, J.-J. Jiang, Y.-Z. Fan, S.-P. Zheng, C.-C. Cao, Y.-H. Li, D. Fenske and C.-Y. Su, Angew. Chem., Int. Ed., 2016, 55, 9932–9936 CrossRef CAS PubMed .
  130. J. Zheng, R. S. Vemuri, L. Estevez, P. K. Koech, T. Varga, D. M. Camaioni, T. A. Blake, B. P. McGrail and R. K. Motkuri, J. Am. Chem. Soc., 2017, 139, 10601–10604 CrossRef CAS PubMed .
  131. S. Chu and A. Majumdar, Nature, 2012, 488, 294 CrossRef CAS PubMed .
  132. T. C. T. Pham, S. Docao, I. C. Hwang, M. K. Song, D. Y. Choi, D. Moon, P. Oleynikov and K. B. Yoon, Energy Environ. Sci., 2016, 9, 1050–1062 CAS .
  133. D. W. Colcleugh and E. A. Moelwyn-Hughes, J. Am. Chem. Soc., 1964, 2542–2545 RSC .
  134. M.-H. Zeng, Q.-X. Wang, Y.-X. Tan, S. Hu, H.-X. Zhao, L.-S. Long and M. Kurmoo, J. Am. Chem. Soc., 2010, 132, 2561–2563 CrossRef CAS PubMed .
  135. D. F. Sava, M. A. Rodriguez, K. W. Chapman, P. J. Chupas, J. A. Greathouse, P. S. Crozier and T. M. Nenoff, J. Am. Chem. Soc., 2011, 133, 12398–12401 CrossRef CAS PubMed .
  136. K. S. Park, Z. Ni, A. P. Côté, J. Y. Choi, R. Huang, F. J. Uribe-Romo, H. K. Chae, M. O’Keeffe and O. M. Yaghi, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 10186–10191 CrossRef CAS PubMed .
  137. K. W. Chapman, D. F. Sava, G. J. Halder, P. J. Chupas and T. M. Nenoff, J. Am. Chem. Soc., 2011, 133, 18583–18585 CrossRef CAS PubMed .
  138. D. F. Sava, K. W. Chapman, M. A. Rodriguez, J. A. Greathouse, P. S. Crozier, H. Zhao, P. J. Chupas and T. M. Nenoff, Chem. Mater., 2013, 25, 2591–2596 CrossRef CAS .
  139. S. S.-Y. Chui, S. M.-F. Lo, J. P. H. Charmant, A. G. Orpen and I. D. Williams, Science, 1999, 283, 1148–1150 CrossRef CAS PubMed .
  140. D. F. Sava, T. J. Garino and T. M. Nenoff, Ind. Eng. Chem. Res., 2012, 51, 614–620 CrossRef .
  141. Z. Wang, Y. Huang, J. Yang, Y. Li, Q. Zhuang and J. Gu, Dalton Trans., 2017, 46, 7412–7420 RSC .
  142. V. Safarifard and A. Morsali, CrystEngComm, 2014, 16, 8660–8663 RSC .
  143. Z.-Q. Jiang, F. Wang and J. Zhang, Inorg. Chem., 2016, 55, 13035–13038 CrossRef CAS PubMed .
  144. S. Yao, X. Sun, B. Liu, R. Krishna, G. Li, Q. Huo and Y. Liu, J. Mater. Chem. A, 2016, 4, 15081–15087 CAS .
  145. S. Parshamoni, S. Sanda, H. S. Jena and S. Konar, Chem. – Asian J., 2015, 10, 653–660 CrossRef CAS PubMed .
  146. R.-X. Yao, X. Cui, X.-X. Jia, F.-Q. Zhang and X.-M. Zhang, Inorg. Chem., 2016, 55, 9270–9275 CrossRef CAS PubMed .
  147. G. Mehlana, G. Ramon and S. A. Bourne, Microporous Mesoporous Mater., 2016, 231, 21–30 CrossRef CAS .
  148. W.-Q. Xu, Y.-H. Li, H.-P. Wang, J.-J. Jiang, D. Fenske and C.-Y. Su, Chem. – Asian J., 2016, 11, 216–220 CrossRef CAS PubMed .
  149. M. S. Deshmukh, A. Chaudhary, P. N. Zolotarev and R. Boomishankar, Inorg. Chem., 2017, 56, 11762–11767 CrossRef CAS PubMed .
  150. H. Kitagawa, H. Ohtsu and M. Kawano, Angew. Chem., Int. Ed., 2013, 52, 12395–12399 CrossRef CAS PubMed .
  151. Z. M. Wang, Y. J. Zhang, T. Liu, M. Kurmoo and S. Gao, Adv. Funct. Mater., 2007, 17, 1523–1536 CrossRef CAS .
  152. Q.-K. Liu, J.-P. Ma and Y.-B. Dong, Chem. Commun., 2011, 47, 7185–7187 RSC .
  153. Y. Rachuri, K. K. Bisht, B. Parmar and E. Suresh, J. Solid State Chem., 2015, 223, 23–31 CrossRef CAS .
  154. N.-X. Zhu, C.-W. Zhao, J.-C. Wang, Y.-A. Li and Y.-B. Dong, Chem. Commun., 2016, 52, 12702–12705 RSC .
  155. Z. Yan, Y. Yuan, Y. Tian, D. Zhang and G. Zhu, Angew. Chem., Int. Ed., 2015, 54, 12733–12737 CrossRef CAS PubMed .
  156. Q.-Q. Dang, X.-M. Wang, Y.-F. Zhan and X.-M. Zhang, Polym. Chem., 2016, 7, 643–647 RSC .
  157. X. Qian, Z.-Q. Zhu, H.-X. Sun, F. Ren, P. Mu, W. Liang, L. Chen and A. Li, ACS Appl. Mater. Interfaces, 2016, 8, 21063–21069 CAS .
  158. E. Stockel, X. Wu, A. Trewin, C. D. Wood, R. Clowes, N. L. Campbell, J. T. A. Jones, Y. Z. Khimyak, D. J. Adams and A. I. Cooper, Chem. Commun., 2009, 212–214 RSC .
  159. Y. Chen, H. Sun, R. Yang, T. Wang, C. Pei, Z. Xiang, Z. Zhu, W. Liang, A. Li and W. Deng, J. Mater. Chem. A, 2015, 3, 87–91 CAS .
  160. A. Sigen, Y. Zhang, Z. Li, H. Xia, M. Xue, X. Liu and Y. Mu, Chem. Commun., 2014, 50, 8495–8498 RSC .
  161. C. Pei, T. Ben, S. Xu and S. Qiu, J. Mater. Chem. A, 2014, 2, 7179–7187 CAS .
  162. C. M. González-García, J. F. González and S. Román, Fuel Process. Technol., 2011, 92, 247–252 CrossRef .
  163. K. W. Chapman, P. J. Chupas and T. M. Nenoff, J. Am. Chem. Soc., 2010, 132, 8897–8899 CrossRef CAS PubMed .
  164. K. Funabashi, T. Fukasawa and M. Kikuchi, Nucl. Technol., 1995, 109, 366–372 CrossRef CAS .
  165. B. Li, X. Dong, H. Wang, D. Ma, K. Tan, Z. Shi, Y. J. Chabal, Y. Han and J. Li, Faraday Discuss., 2017, 201, 47–61 RSC .
  166. D. Banerjee, C. M. Simon, A. M. Plonka, R. K. Motkuri, J. Liu, X. Chen, B. Smit, J. B. Parise, M. Haranczyk and P. K. Thallapally, Nat. Commun., 2016, 7, ncomms11831 CrossRef CAS PubMed .
  167. D. Banerjee, A. J. Cairns, J. Liu, R. K. Motkuri, S. K. Nune, C. A. Fernandez, R. Krishna, D. M. Strachan and P. K. Thallapally, Acc. Chem. Res., 2015, 48, 211–219 CrossRef CAS PubMed .
  168. H. Wang, K. Yao, Z. Zhang, J. Jagiello, Q. Gong, Y. Han and J. Li, Chem. Sci., 2014, 5, 620–624 RSC .
  169. X. Chen, A. M. Plonka, D. Banerjee, R. Krishna, H. T. Schaef, S. Ghose, P. K. Thallapally and J. B. Parise, J. Am. Chem. Soc., 2015, 137, 7007–7010 CrossRef CAS PubMed .
  170. J. Liu, P. K. Thallapally and D. Strachan, Langmuir, 2012, 28, 11584–11589 CrossRef CAS PubMed .
  171. O. Shekhah, Y. Belmabkhout, Z. Chen, V. Guillerm, A. Cairns, K. Adil and M. Eddaoudi, Nat. Commun., 2014, 5, 4228 CAS .
  172. E. Barea, C. Montoro and J. A. R. Navarro, Chem. Soc. Rev., 2014, 43, 5419–5430 RSC .
  173. I. Ahmed and S. H. Jhung, J. Hazard. Mater., 2016, 301, 259–276 CrossRef CAS PubMed .
  174. J. G. Vitillo, RSC Adv., 2015, 5, 36192–36239 RSC .
  175. K. S. Lackner, Science, 2003, 300, 1677–1678 CrossRef CAS PubMed .
  176. T. M. McDonald, W. R. Lee, J. A. Mason, B. M. Wiers, C. S. Hong and J. R. Long, J. Am. Chem. Soc., 2012, 134, 7056–7065 CrossRef CAS PubMed .
  177. P. Nugent, Y. Belmabkhout, S. D. Burd, A. J. Cairns, R. Luebke, K. Forrest, T. Pham, S. Ma, B. Space, L. Wojtas, M. Eddaoudi and M. J. Zaworotko, Nature, 2013, 495, 80–84 CrossRef CAS PubMed .
  178. J. Yu, L.-H. Xie, J.-R. Li, Y. Ma, J. M. Seminario and P. B. Balbuena, Chem. Rev., 2017, 117, 9674–9754 CrossRef CAS PubMed .
  179. H. B. Tanh Jeazet, C. Staudt and C. Janiak, Dalton Trans., 2012, 41, 14003–14027 RSC .
  180. E. D. Bloch, M. R. Hudson, J. A. Mason, S. Chavan, V. Crocellà, J. D. Howe, K. Lee, A. L. Dzubak, W. L. Queen, J. M. Zadrozny, S. J. Geier, L.-C. Lin, L. Gagliardi, B. Smit, J. B. Neaton, S. Bordiga, C. M. Brown and J. R. Long, J. Am. Chem. Soc., 2014, 136, 10752–10761 CrossRef CAS PubMed .
  181. Q. Min Wang, D. Shen, M. Bülow, M. Ling Lau, S. Deng, F. R. Fitch, N. O. Lemcoff and J. Semanscin, Microporous Mesoporous Mater., 2002, 55, 217–230 CrossRef .
  182. A. Aijaz, A. Karkamkar, Y. J. Choi, N. Tsumori, E. Rönnebro, T. Autrey, H. Shioyama and Q. Xu, J. Am. Chem. Soc., 2012, 134, 13926–13929 CrossRef CAS PubMed .
  183. M. S. Shah, M. Tsapatsis and J. I. Siepmann, Chem. Rev., 2017, 117, 9755–9803 CrossRef CAS PubMed .
  184. L. Hamon, C. Serre, T. Devic, T. Loiseau, F. Millange, G. Férey and G. D. Weireld, J. Am. Chem. Soc., 2009, 131, 8775–8777 CrossRef CAS PubMed .
  185. Y. Belmabkhout, R. S. Pillai, D. Alezi, O. Shekhah, P. M. Bhatt, Z. Chen, K. Adil, S. Vaesen, G. De Weireld, M. Pang, M. Suetin, A. J. Cairns, V. Solovyeva, A. Shkurenko, O. El Tall, G. Maurin and M. Eddaoudi, J. Mater. Chem. A, 2017, 5, 3293–3303 CAS .
  186. X. Cui, Q. Yang, L. Yang, R. Krishna, Z. Zhang, Z. Bao, H. Wu, Q. Ren, W. Zhou, B. Chen and H. Xing, Adv. Mater., 2017, 29, 1606929 CrossRef PubMed .
  187. K. Tan, S. Zuluaga, H. Wang, P. Canepa, K. Soliman, J. Cure, J. Li, T. Thonhauser and Y. J. Chabal, Chem. Mater., 2017, 29, 4227–4235 CrossRef CAS .
  188. A. M. Ebrahim, B. Levasseur and T. J. Bandosz, Langmuir, 2013, 29, 168–174 CrossRef CAS PubMed .
  189. G. W. Peterson, J. J. Mahle, J. B. DeCoste, W. O. Gordon and J. A. Rossin, Angew. Chem., Int. Ed., 2016, 55, 6235–6238 CrossRef CAS PubMed .
  190. J. E. Mondloch, M. J. Katz, W. C. Isley Iii, P. Ghosh, P. Liao, W. Bury, G. W. Wagner, M. G. Hall, J. B. DeCoste, G. W. Peterson, R. Q. Snurr, C. J. Cramer, J. T. Hupp and O. K. Farha, Nat. Mater., 2015, 14, 512 CrossRef CAS PubMed .


These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2018