Structural stability of metal organic frameworks in aqueous media – Controlling factors and methods to improve hydrostability and hydrothermal cyclic stability

Accepted Manuscript Review Structural stability of metal organic frameworks in aqueous media - Controlling factors and methods to improve hydrostability and hydrothermal cyclic stability Najam ul Qadir, Syed A.M. Said, Haitham M. Bahaidarah PII: DOI: Reference:

S1387-1811(14)00537-X http://dx.doi.org/10.1016/j.micromeso.2014.09.034 MICMAT 6779

To appear in:

Microporous and Mesoporous Materials

Received Date: Revised Date: Accepted Date:

26 May 2014 3 September 2014 9 September 2014

Please cite this article as: N.u. Qadir, S.A.M. Said, H.M. Bahaidarah, Structural stability of metal organic frameworks in aqueous media - Controlling factors and methods to improve hydrostability and hydrothermal cyclic stability, Microporous and Mesoporous Materials (2014), doi: http://dx.doi.org/10.1016/j.micromeso.2014.09.034

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

Structural Stability of Metal Organic Frameworks in Aqueous Media – Controlling Factors and Methods to Improve Hydrostability and Hydrothermal Cyclic Stability Najam ul Qadir*, Syed A. M. Said and Haitham M. Bahaidarah Center of Research Excellence in Renewable Energy, King Fahd University of Petroleum and Minerals, Dhahran-31261, KSA

ABSTRACT Metal organic frameworks (MOFs) have recently emerged as a center of attention amongst the class of more traditional porous materials including zeolites, activated carbons, and silica gels, due to their significantly larger pore volumes and tunability of pore geometry. A large amount of literature exists incorporating the gas adsorption and separation characteristics of MOFs; however, only a limited portion has been dedicated to the water adsorption properties of these novel materials. Water adsorption capacities of MOFs are known to substantially exceed those of traditional porous materials, due to which they are continuously gaining increasing attention for potential applications in water desalination and purification, dehumidification, and adsorption cooling technologies. However, a vast majority of these materials is characterized by structural degradation on either immediate or prolonged exposure to moist environments, resulting in partial or complete deterioration of water adsorption characteristics of the framework. This article reviews the various structural parameters which control the dimensional stability of MOFs in aqueous media, as well as different strategies which have been reported in literature for improving the hydrothermal resistance of these materials for potential applications in the industrial sector. Keywords: structural; frameworks; hydrothermal; cyclic; hydrostability *

Corresponding author (N.U. Qadir: [email protected]) 1

1. Introduction Metal Organic Frameworks (MOFs) are categorized as unique functional materials with exceptional features such as extremely large surface area to volume ratio, and a tremendous structural flexibility, which leads to the degree of geometric and chemical variability not exhibited by any other porous material. The structure of MOFs primarily consists of two major components – metal centers acting as joints in the structure, and organic linkers which act as struts and bridge the neighboring joints. The vast proportion of research based on the development of different types of MOF architectures has been devoted to the potential usefulness of these materials in applications like storage of light gases [1–4], gas separation [5–10], catalysis [11–15], optics [16–18], magnetism [19–21], and others [2,22–24]. A relatively negligible amount of research effort has been dedicated so far concerning the water adsorption characteristics of these materials, and the relevant applications where this unique attribute of MOFs is of fruitful importance. However, a majority of MOFs have been known to lose structural integrity in an aqueous medium (ambient moisture/water vapor, room temperature water, boiling water, steam, aqueous acidic/basic solutions etc.), which has been considered as a major constraint regarding the potential usefulness of these materials in applications like adsorption cooling [25–27], and water desalination [28–30]. This is particularly true for MOFs based on divalent metal cations combined with organocarboxylate bridging ligands [31], which can be subject to hydrolysis and thermal decomposition on exposure to an aqueous medium [39,52]. On exposure to an aqueous medium, water molecules have been reported to attack the metal connectors within MOFs, displacing ligands and causing phase changes, loss in crystallinity, and/or structural decomposition resulting in significant loss in specific surface area and hence total pore volume of the materials [52,58]. For example, the [Zn4O]6+ clusters present 2

in MOF-5 are easily hydrolyzed by water vapor, forming a non-porous product containing zinc(II) hydroxide chains commonly referred to as MOF-69C [31,38,40]. Hence a water-stable or a hydrostable MOF can be characterized by a structure which can successfully block the intrusion of water molecules into the framework, thereby effectively shielding the metal centers and ligands in order to avoid any potential phase changes, and associated losses in crystallinity and overall pore volume of the structure. Isoreticular MOFs (IRMOFs) incorporate oxide-centered Zn4O tetrahedral clusters as metal centers connected by dicarboxylate organic linkers, forming a three-dimensional (3-D) connected porous framework.

IRMOF-1, also known as MOF-5 (Zn-BDC, BDC = 1,4-

benzenedicarboxylate), was the first known member of this family and is an archetypal MOF which is characterized by the highest degree of instability in an aqueous medium of all the MOFs known to date. IRMOFs in general, and MOF-5 in particular, are known to be highly sensitive to moisture because it is believed that the weak zinc-oxygen coordination allows for an attack by water molecules [31,40]. The water molecules cause the framework to decompose, resulting in lower specific surface area and thus poor adsorption properties. A variety of research approaches have been adopted in a number of recent publications in order to increase the structural stability of MOF-5 framework in an aqueous environment. These mainly include the functionalization of the BDC linker used in the MOF-5 framework with one or more hydrophobic functionalities during in-situ MOF synthesis [42], post-synthetic modification of the BDC linker with hydrophobic moieties like alkyl chains [43], coating the MOF-5 structure with a more hydrophobic material [172], incorporation of second-phase hydrophobic nanoparticles inside the MOF-5 framework [174], partial exchange of Zn(II) ions in the MOF-5 framework with other transition metal ions such as Ni(II) [178], and confinement of MOF-5 particles in a water3

resistant matrix [181]. Table 1 summarizes all these different approaches adopted so far for improving the structural stability of MOF-5 in an aqueous medium, and the type(s) of aqueous medium/media (room temperature water, boiling water, ambient moisture, steam, aqueous acidic solutions, aqueous basic solutions) for which they have been proven practically successful. It can be easily seen in Table 1 that the approaches utilized so far in improving the stability of MOF-5 in aqueous media have only been successful either for ambient moisture or water at room temperature. Hence, there exists a need to modify these approaches or design better ones in order to stabilize the MOF-5 structure in more harsh humidity conditions like boiling water, steam, and aqueous acidic and basic solutions. MOFs containing carboxylate-metal bonds have been reported in literature to exhibit varying degrees of hydrothermal stability. Huang et al. [32] reported that MOF-5 exhibits a decrease in its BET surface area from 900 to 45 m2g-1 upon water exposure, eventually reducing to a new crystal structure. They suggested that Zn4O clusters in the framework undergo hydrolysis, on interaction with water molecules, to form the Zn ion and the corresponding organic linker acid (benzenedicarboxylic acid for MOF-5, and 1,3,5-benzenetribenzoate (BTB) for MOF-177) [32]. However, they proposed that the hydrolysis reaction might also occur favorably on framework exposure to an acidic medium. Furthermore, DMOFs have been experimented to prove that water stability in MOFs can also be improved through the incorporation of pendant functional groups, as well as replacement of the metal center in the secondary building unit (SBU) [33–35]. On the contrary, MIL-53 (MIL = Material of Institut Lavosier), an aluminum MOF with benzene dicarboxylate linkers, has been observed to reversibly change its conformation upon hydration, without undergoing permanent structural

4

changes [36]. Surble et al. [37] also designed the MIL-88 analogs, which was a series of waterstable Fe(III) and Cr(III) MOFs containing various dicarboxylate linkers. The water sensitivity of these materials has been a subject of increasing attention amongst the MOF research community dedicated to the development of water-stable frameworks, both at ambient and elevated temperatures [32,39–40,187–193]. The main reason suggested for structural degradation in an aqueous medium, is the binding of water molecules to unsaturated metal sites [41], with the exception of MOF-5 which has no coordinatively unsaturated sites but still undergoes structural disintegration on exposure to water vapor [40]. Consequently, many research attempts have successfully resulted in finding ways to develop water stable MOFs using functionalization with hydrophobic functional groups, either by direct synthesis using functionalized linkers [39,42], or by post-synthetic modification of the framework [43]. It has been generally observed that MOFs which are hydrothermally synthesized show higher water stability than those which are synthesized in the absence of an aqueous medium. The effect of aqueous media on the structural stability of MOFs has been studied both experimentally and theoretically [41,44–49]. Sabo et al. [50] experimentally investigated the step-wise hydrolysis of MOF-5, which finally leads to the complete disintegration of the framework. They proposed that the metal-ligand bond disintegrates during this hydrolysis reaction, and the water molecule dissociates into a hydroxide anion which bonds to the metal center, and a proton which binds to the displaced ligand. Low et al. [52] used quantum mechanical calculations of MOF cluster models followed by experimental validation, to conclude that hydrothermal stability of MOFs is basically dependent on the metal-linker bond strength. It has further been suggested that structural stability in an aqueous medium is not only 5

governed by the metal-ligand bond strength, but also by factors like the type of selected metal center, its coordination state, and framework dimensionality [53]. Greathouse and Allendorf [38] utilized a constant-pressure force-field based molecular dynamics simulation to predict the interaction of the MOF-5 framework with water molecules, as well as the effect of their weight fraction on the mechanical stability of the framework. On the basis of the possible reaction for MOF-5 proposed by Greathouse and Allendorf, Li and Yang [51] suggested a hydrolysis reaction for MOF-177:

Zn 4 O(BTB)2 + 4H 2 O → [(Zn 4 O )(H 2 O )4 (BTB)] + BTB3− 3+

(1)

It was proposed that the following reaction may also occur simultaneously with the above reaction [51]:

(Zn 4 O )(H 2 O)4 (BTB)3+ → Zn(OH )2 + 3ZnO + BTB(3H ) + 3H +

(2)

The inner surface of porous IRMOFs mainly exhibits non-polar functional groups together with Zn ions that are coordinated to oxygen atoms. This surface has an overall hydrophobic character, while the Zn4O clusters included the framework have been proposed to act as “weak hydrophilic defects” [54–55]. When only one water molecule is present in the pore, the bound state is frequently visited but it is less stable than the water-free state, and no chemisorption of any kind is taking place. However at a higher water loading, a water cluster can be formed at the Zn cluster site which has been shown to stabilize the water-bound state, which rapidly transforms into a linker-displaced state, where water has fully displaced one arm of a linker and corresponds to the loss of the material’s fully ordered structure [56]. Thus an overall hydrophobic MOF material can also lose stability when exposed to an aqueous medium – an anomaly that has not been fully understood until now. In an ab initio molecular dynamics (MD) 6

study, Marta et al. [56] have described the possible mechanisms involved in hydration of IRMOFs in an aqueous medium, particularly addressing the apparent instability of MOF-5 despite the existence of its hydrophobic framework. Since MOF-5 bears no coordinatively unsaturated sites in its framework, its structural degradation in the presence of water vapor is highly intriguing. In order to investigate the effect of hydration level on the stability of IRMOFs, a loading of four water molecules per unit cell of the IRMOF framework was utilized in the MD simulations. Figure 1 shows the intermediate hydration states of a Zn2+ coordination sphere in the structure of IRMOF-0h (IRMOF framework with zero or one bound water molecule) [56]. The cation has a tetrahedral environment in the non-hydrated material as shown by state A in Figure 1, with four oxygen atoms coordinated to it: three from the carboxylate linkers, and one being the central O atom of the Zn4O cluster. When the water molecule binds to the Zn2+ cation, none of these atoms exit the coordination shell, and the Zn2+ thus has a trigonal bipyramidal environment as shown by state B in Figure 1 [56]. It was further confirmed that this state is indeed a stable state, and is also similar in geometry to the hydrated state hypothesized by Low et al. [52]. The overall hydration mechanism which can be deduced from Figure 1 can thus be considered to proceed in two successive steps – (a) water binding to the metallic cluster, and (b) linker displacement, where one water molecule actually replaces on of the linker arms. These two steps are believed to take place rapidly since they could be observed on the timescale of ab initio MD, provided there is sufficient water present in the porous framework [56]. One water molecule has been hypothesized to bind to the metallic cluster, while the extra molecules stabilize both the hydrated and the linker-displaced states so that the overall energy barrier for the hydration reactions becomes of the order of or less than kT, where k represents the Boltzmann’s constant and T denotes the temperature [56]. From the results of their force-field

7

based simulations, Greathouse and Allendorf concluded that water displaced rather indifferently two types of oxygen atoms initially coordinated to the Zn2+ ions – the central atom of the Zn4O tetrahedron, and those of the linkers’ carboxylate groups [38]. This contrasts with the approach of Marta et al. in which two possible hydrated states were characterized at low to medium water loading, none of which displays a rupture (or even a marked elongation) of the central Zn–O bond [56]. They further proposed that this bond is the stronger of the two coordinative bonds, and the Zn2+ cation will prefer keeping a coordination number of five with a trigonal bipyramid geometry, rather than break the central Zn–O bond. This difference with the findings of Greathouse and Allendorf was suggested to be due to a bias of the classical force field, which has a tendency to describe the Zn2+ as tetrahedrally coordinated [56]. Recent studies concerning the development of water-stable MOFs have concluded that MOFs based on Al-carboxylate coordination chemistry are amongst the most hydrothermally stable of these materials reported to date [52,57]. Among the carboxylate-based MOFs, hydrophilic MIL-100 and MIL-101, 8-coordinated Zr-MOF UiO-66, and MOF-74/CPO-27 have all been described as water-stable, but for certain prescribed levels of environmental humidity [44,58–59]. Similarly, nitrogen-coordinated MOFs have been observed to be fairly water-stable due to the higher basicity of these ligands compared to carboxylic acid [52,60]. For example imidazolate-based ZIF-8 (ZIF = zeolitic imidazolate framework) has been investigated to retain structural stability upon exposure to high levels of humidity [44]; however, loss of framework integrity has been observed upon dispersion in water for 3 months [61]. Moreover, this compound does not adsorb water even to high relative pressures [62]. The water adsorption isotherm of ZIF-8 has also been observed to indicate its strong hydrophobic character, showing only insignificant water adsorption up to p/p0 = 0.6 (sample pressure/saturation vapor pressure) 8

followed by a slight increase of approximately 10 cm3g-1 up to p/p0 = 0.8 and a final steep increase due to water condensation to about 150 cm3g-1 in the higher pressure region [62]. In general, the hydrothermal stability of MOFs containing imidazolates and other nitrogencontaining ligands has been observed to be higher than that of carboxylate-based MOF materials [52]. It has also been proposed that structural stability of MOFs in aqueous media can be improved via chemical attachment of functional moieties, e.g. alkyl or fluorinated groups, resulting in increased hydrophobicity of the framework [42,63–64]. Structural stability of MOFs in aqueous media also depends significantly on steric factors, apart from the metal-ligand acid-base effects discussed above. Ma et al. [60] investigated the shift in hydrothermal stability of three MOF materials with pillared ligands, synthesized using Zn, BDC, and 4,4′ -bipyridine (BPY) functionalized with methyl groups at various positions. They observed that functionalization with methyl groups only at the 2,2′ -position on BPY leads to a structure that retains its structural integrity on exposure to humid air. The

resulting MOF, SCUTC-18, was observed to absorb only 4 wt% of water vapor; however, the relative humidity (RH) at equilibrium was not reported. However, it remains doubtful whether the SCUTC-18 framework appears water-stable owing to the methyl groups shielding the coordination sites, or as a consequence of its limited affinity towards water molecules. However, it is still difficult to completely suppress adsorption of water molecules in these hydrophobic frameworks, especially when the exposure time exceeds a critical duration. In other words, if the higher observed stability is solely attributed to the decreased uptake of water molecules, then the quantity of water adsorbed over long exposure times might eventually disrupt the framework significantly [33]. Hence, there is an ever-increasing need for the development of water-stable MOFs which can withstand adsorbed moisture over prolonged exposure times, in order to be 9

considered commercially feasible for applications necessitating large amounts of water adsorption. However, the effects of porosity, topology, ligand character, and metal-ligand coordination must be decoupled before a comprehensive investigation of the mechanisms responsible for structural degradation of MOFs in humid environments can be conducted. Moreover, a systematic approach that is required to conduct such an elaborative study is difficult to design owing to the limited available variety of isostructural families of MOFs with a sufficiently broad range of functional groups. Recently, Jérôme et al. [65] presented a comprehensive review article based on structural stability of MOFs in aqueous media including pure water and aqueous acidic/basic solutions, degradation mechanisms of MOFs in these aqueous media, and mechanisms responsible for water adsorption/desorption mechanisms of these materials. The article concluded with a review of the various applications which involve water adsorption by MOFs including heat pumps and chillers, proton conducting membranes, and adsorption chillers. However, Jérôme et al. did not review the potential methodologies that have been adopted so far by various research groups to improve the structural stability of MOFs in aqueous media for achieving better usefulness of these materials in various applications they have cited in their article. The current review article basically incorporates two topics considered fundamental in MOF research concerning waterstability of these materials. The first topic further elaborates the various controlling factors presented by Jérôme et al. [65] which govern the structural stability of MOFs in aqueous media including ambient moisture, liquid water at room temperature, boiling water, steam at different levels of temperature and relative humidity, and aqueous acidic/basic solutions. The second topic presents a review of the various methodologies that have been adopted so far by various research groups with an aim to improve the structural stability of MOFs in aqueous media for better 10

utilization of these materials in continuously emerging potential applications on a commercial scale. It is not expected to incorporate all the different strategies that have been proposed so far for improving the water-stability of these materials; however, it still entails the key approaches that have been proposed recently, and can prove beneficial concerning the selection of the right approach given the chemistry and dimensionality of the framework at hand. The second section includes the description of the various physical and chemical entities which directly influence the stability of MOFs in humid air or in liquid water, both at ambient and at elevated temperatures. The third section describes a variety of methods which have recently been adopted to enhance the hydrostability of MOFs for potential usage in applications where framework exposure to aqueous media cannot be avoided. The fourth section introduces the concept of hydrothermal cyclic stability, supplemented by a few newly published works involving methods to enhance this property in MOFs, for intended use as potential adsorbents in commercial heat transformation applications. Figure 2 summarizes the various factors controlling the structural stability of MOFs in aqueous media, as well as methods which can be adopted to improve the hydrostability and hydrothermal cyclic stability of the framework. 2. Factors controlling structural stability of MOFs in aqueous media

This section describes the different physical and chemical parameters which have been observed to influence the structural stability of MOFs in an aqueous medium, which represents ambient moisture at varying levels of humidity, liquid water at both ambient and elevated temperatures, steam, and aqueous acidic and basic solutions.

11

2.1

Metal-ligand bond strength

It has been suggested that metal-ligand bonds in MOF structures are either based on a Lewis-acid–Lewis-base interaction of a localized lone electron pair, or on a non-specific electrostatic interaction [66]. In general, metal centers with lower coordination numbers are expected to result in shorter bond lengths, and have been suggested to result in relatively greater metal-ligand bond strengths [67]. Metal-ligand bond strengths in MOFs containing divalent metal centers are observed to decrease in the following order: Fe-O (468 kJ mol-1) > Cr-O (374 kJ mol-1) > Cu-O (372 kJ mol-1) > Zn-O (365 kJ mol-1) [52]. Based on this ordering, HKUST-1 has been proposed to be more water-stable than MOF-5, and MIL-101 has been predicted to be slightly more stable than HKUST-1, due to the presence of trivalent Cr (Cr2O3 at 465 kJ mol-1) [52]. However, these predictions still cannot be considered practically feasible, since they do not incorporate the effects of the overall oxidation state of the SBUs on their resulting structural stability in an aqueous environment. In reality, metal-ligand bond strength cannot be considered as a completely independent factor affecting the water-stability of the framework, since other factors like basicity of the ligand, and coordination number of the metal center directly affect the strength of the metal-ligand bond, which eventually influences the resultant structural stability in an aqueous medium. 2.2

Basicity of the ligand

It has been proposed that more basic the ligand, the stronger will be the metal-ligand bond [67]. This is expected to order the metal-ligand bond strengths as carboxylate (pKa ≈ 5 ) ≈

pyridine (pKa = 5) << O2- (pKa ≈ 36 ) [68–69]. The greater stability of pillared MOFs has been attributed to the higher basicity, i.e., higher pKa value, of the pillar ligand (e.g. DABCO, BPY,

12

DABCO = 1,4-diazabicyclo[2.2.2]octane, BPY = 4, 4 ′ -bipyridyl) relative to the dicarboxylate ligand (e.g. BDC, TMBDC, TMBDC = tetramethyl-1,4-benzenedicarboxylate) as summarized in Table 2 [73], and shown in Figure 3 [73]. Based on the same logic, MOFs based on the highly basic pyrazole (pKa ≈ 19.8 ) [70] and imidazole (pKa ≈ 18.6 ) [70] ligands are found to be stable upon exposure to humid conditions [71–72]. This reasoning is further validated by the work of Low et al. [52], who found that the strength of the bond between the metal oxide cluster and the bridging linker bears a significant contribution towards the overall hydrothermal stability of the framework; the metal, being a Lewis acid, is expected to form a stronger bond with the more basic (higher pKa) ligand. However, in case of MOF-508, it has been observed that it is hydrothermally stable (after 90% RH exposure) when synthesized with a pillar ligand (BPY) of lesser basicity (pKa = 4.6), compared to DABCO (pKa = 8.86) [73]. Since the Lewis acid metal sites have been suggested to form stronger bonds with the more basic ligands [52], DMOF is thus expected to be more hydrothermally stable than MOF-508. However, in reality MOF-508 has been proposed to be more water-stable than DMOF due to the presence of a two-fold catenated framework [73]. In contrast, DMOF has been observed more hydrothermally stable than MOF-508-TM even under harsh humidity conditions, due to a combination of shielded zinc clusters and higher basicity of the DABCO ligand [73]. It has already been shown that pyrazolate- and imidazolate-based MOFs, in general, exhibit higher chemical resistivity towards water molecules than carboxylate-based MOFs [31,51–52,70–71]. This observation has been correlated with the relatively higher basicity (higher pKa) of the nitrogen-based ligand than the oxygen-based carboxylate ligand, and consequently a higher expected metal–ligand bond strength. However, a few recent studies reveal that the carboxylate and pyrazolate functionalities can also be combined within the same 13

ligand for achievement of desired framework characteristics like preferential sensing of aromatic fluids [76–77]. Although carboxylate-based MOFs are generally characterized by lower structural stability in aqueous media than pyrazolate- and imidazolate-based frameworks, a few exceptions like MIL-53(Al,Cr), MIL-101(Cr), MIL-100(Cr), UiO-66(Zr), and MIL-125(Ti), have been reported to exhibit fairly high levels of hydrothermal stability even at elevated temperatures [36,58,78–80]. These exceptions suggest that structural stability of MOFs in an aqueous medium is not only governed by the basicity of the ligand and consequently the strength of the metal– ligand bond, but also depends upon factors such as type of metal center/cluster, its coordination state, and framework dimensionality [52,57,60]. 2.3

Coordination number and type of metal center

It has been reported that MOFs containing 6-coordinate (usually octahedral) metal ions show higher hydrothermal stability than those containing 4-coordinate (usually tetrahedral) metal ions [52]. One probable justification to this observed trend is the greater amount of space available to a water molecule to interact with a tetrahedrally coordinated metal than with a metal with octahedral coordination, thus resulting in a relatively lower energy barrier for linker displacement. For example, MOF-5 framework has been observed to be less stable in an aqueous medium than Zn-MOF-74 [81]. Cu-BTC is comprised of 1,3,5-benzene tricarboxylate (BTC) organic linkers bound by a dicopper tetracarboxylate paddlewheel SBU [82]. Each SBU contains two Cu ions bonded to four tricarboxylate linkers, giving rise to the structure shown in Figure 4 [83]. It has been observed that techniques such as plasma enhanced chemical vapor deposition of perfluroalkanes on Cu-BTC have been utilized to enhance its hydrothermal stability [84]. Both Cu-BTC and Mg-

14

MOF-74 have been reported to possess coordinatively unsaturated metal sites which act as active centers for interaction with water and small molecules with free electrons, making them ideal for the formation of complexes with small Lewis bases [83]. MOF-74 analogs have 1-dimensional (1-D) hexagonal channels approximately 1.1 nm in diameter, surrounded by six helical M–O–C rods (M = metal center), of composition [M2O2](CO2)2, constructed of 5-coordinated M(II) ions as shown in Figure 4 [83]. M–O–C rods in MOF-74 are formed through coordination of the carboxyl and oxy groups of 2,5-dioxoterepthalate with M(II) ions (where M = Zn, Ni, Co, Fe, Mg, Mn, Ca, or Sr) [81]. Mg-MOF-74 has been of particular interest because of its extraordinary uptake of CO2 and SO2 [85–88]. However, it has been observed to degrade even under mild humidity conditions, with a loss of approximately 90% in CO2 uptake capacity upon exposure to 70% RH at 298 K [89]. The stability of MOF-74 analogs in an aqueous medium has been observed to vary with the type of metal center included in the SBU, and can be ranked as: Co > Ni > Zn ≈ Mg [89–90]. Schoenecker et al. [75] observed that Mg-MOF-74 loses 83% of its BET surface area after running water adsorption isotherms up to 90% RH at 298 K; however, they also reported significant retention of the crystal structure via Powder X-ray diffraction (PXRD). UiO-66, a zirconium-based MOF, has been reported to be both thermally and chemically stable [58]. The UiO series of MOFs is defined by the Zr6O4(OH)4 SBU, which is twelvecoordinated to terephthalate ligands in UiO-66 as shown in Figure 4. It has been reported by Lillerud and coworkers [91–92] that the SBU is the key to the exceptional hydrothermal stability of the UiO-series, and has the ability to reversibly change between a hydroxylated [Zr6O4(OH)4] and a dehydroxylated [Zr6O6] structure. However, it has also been reported that many of the UiO analogs, including UiO-67 (biphenyl-4, 4 ′ -dicarboxylate linker), are not as stable to aqueous conditions as UiO-66 [93–94]. In fact, Cohen and coworkers have recently found that UiO-66 15

analogs can be post-synthetically modified via ligand or cation exchange, exhibiting that even UiO-66 may not be as chemically inert as first reported [95–97]. The superior water stability of UiO-66 when compared to Cu-BTC and Mg-MOF-74 can be attributed to various steric and chemical factors. UiO-66 has been observed to possess much narrower pores than the other two MOFs since each SBU is coordinated to twelve organic linkers, making the metal–carboxylate sites less accessible to water molecules. Furthermore, each Zr atom in UiO-66 has a coordination number of 8, making the sites fully saturated and unable to coordinate with water or other guest molecules. Conversely, the metal atom centers in Cu-BTC and Mg-MOF-74 have coordination numbers of 4 and 5 respectively, leaving free metal sites facing the main pore of each framework. These sites chemisorb water and possess the ability to host water clustering even at low RH values, which is further supported by the water isotherms of these materials [83]. Tan et al. [34] studied the interaction of water molecules with one prototypical MOF, M(BDC)(TED)0.5 (M = Cu, Zn, Ni, Co, BDC = 1,4-benzenedicarboxylate, TED = triethylenediamine), containing SBUs of two 5-coordinate Cu cations bridged in a paddle-wheel type configuration [98–105]. M(BDC)(TED)0.5 has been shown to be an excellent absorbent for a variety of gases including H2, CH2, CH4, and other hydrocarbons [103–105]. In-situ infrared (IR) spectroscopy, ex-situ Raman spectroscopy, and PXRD were utilized to investigate the nature of interaction of water molecules with the as-synthesized prototypical MOFs [34]. In order to study the water adsorption behavior of each of the selected frameworks, pressure dependence measurement of D2O vapor adsorption into M(BDC)(TED)0.5 was performed. The experimental results revealed that the type of central metal ions included in the structure is the most contributing factor towards the initial decomposition pathway, and the structural stability of each of the studied frameworks on exposure to D2O vapors. In case of Cu(BDC)(TED)0.5, a hydrolysis 16

reaction of the Cu-O-C group was observed to induce the paddle-wheel structural decomposition [34]. However, for the Zn(BDC)(TED)0.5 and Co(BDC)(TED)0.5 frameworks, the water molecules were observed to gradually replace the TED pillars, and bond to the apical sites of the paddle-wheels Zn2(COO)4 and Co2(COO)4. It was deduced from the study that the overall hydrothermal stability of the isostructural MOFs, M(BDC)(TED)0.5, follows the order: Cu-MOF < Ni-MOF > Zn-MOF > Co-MOF, which was found to be correlated with the bond dissociation energy of the diatomic metal-oxygen complexes, and the overall stability constants of the metalamine complexes [34]. 2.4

Oxidation state of the metal center

It is generally expected that higher the oxidation state of the metal center or the metal cluster, the higher will be the relative stability of the framework towards water molecules [52]. A proposed strategy of comparing clusters to individual metals is to take the charge of the cluster, and divide it by the total number of metals included in the cluster [52]. For example, the relative positive charge on a single Zn atom in the Zn4O6+ cluster in MOF-5 is 1.5, while the relative positive charges on a Cu atom in HKUST-1, and a Cr atom in MIL-101, are 2.0 in each case. Qualitatively, this suggests that the SBUs in HKUST-1 and MIL-101 should form more stable bonds with their respective linkers, than those formed between the MOF-5 SBU and the BDC linkers. Based on a similar logic, it has been proposed that individual Zn atoms, each of which possesses a 2+ oxidation state in ZIF-8, are expected to form stronger metal-ligand bonds than those observed to form between the Zn4O6+ SBU and the six BDC linkers [52]. Hence, it can also be assumed that ZIF-8 is more chemically resistant than MOF-5 towards hydration reactions of the type expressed in equations (1) and (2). However, it is worth mentioning that this theoretical assumption fails to incorporate the obvious difference between the imidazolate nitrogen atom17

metal, and the carboxylate oxygen-metal bond strengths. As discussed earlier, the Zn-N bonds in ZIF-8 are expected to be stronger than the Zn-O bonds in MOF-5. It has been observed that M3+-containing SBUs in MOFs are more resistant towards moisture as compared to M2+-containing SBUs [52]. Cr3+ containing MIL-53-Cr has been observed to exhibit structural stability towards reaction with water vapor, which is comparable to that observed for MIL-101-Cr [78], as shown in Table 3 [52]. The MIL-53-SBU is composed of infinite chains of corner-sharing M3+octahedra bridged by the BDC linker, to form 1-D lozengeshaped pores throughout the framework, as shown in Figure 5(a) [52]. The Cr–O bonds between the ligand and metal centers in the 1-D chains of MIL-53-Cr are almost equally resistant to reaction with water, as those of the Cr3-µ-OF6+ trimer in MIL-101-Cr. This indicates that the relative strength of the M–O bond is a more contributing factor towards structural stability as compared to coordination geometry, or valence, for the M3+-containing MOFs than for the M2+containing ones. This hypothesis has been confirmed by comparing the stability evaluated for isostructural MIL-53-Al to that of MIL-53-Cr [36], i.e., the Al–O bond in alumina is evaluated to be stronger than the Cr–O bond in Cr(III) oxide (514 kJmol-1 compared to 447 kJmol-1). It follows that the computed value of the activation energy for displacement of the BDC linker by water molecules, is higher for MIL-53-Al, than that evaluated for MIL-53-Cr (43.5 kcalmol-1 versus 30.4 kcalmol-1) as shown in Table 3 [52]. The higher stability towards reaction with water vapor experimentally observed for MIL-53-Al, than MIL-101-Cr, further supports this hypothesis that Al3+–containing MOFs are generally more hydrothermally stable than Cr3+– containing frameworks. Such reasoning can be further extended to other Al-containing MOFs such as MIL-53 and MIL-110 [106]. The SBUs of Al-MIL-53 and MIL-110 are shown schematically in Figure 5 [52]. The metal cations in MIL-110 are located in octamer SBUs of Al18

octahedra, linked through BTC linkers to form 1D channels as shown in Figure 5(b). There are two trimers of edge-sharing octahedral, and two edge-sharing capping octahedra octamer SBU. Formally, the caps are observed to be AlO3(OH)3, and the trimers to be AlO2(OH)3(H2O), or AlO2(OH)4, depending upon the termination species present. MIL-110 has been determined experimentally to show hydrothermal stability comparable to that of MIL-53-Al [107], which is further supported by the data shown in Table 3 [52]. 2.5

Chemical functionality of the linker

As bridging ligands, carboxylates are of immense interest in the construction of MOFs. However, one of the major obstacles towards the industrial commercialization of water-stable MOFs is the instability of a majority of carboxylate-based MOFs on exposure to liquid water, or even high humidity. In particular, IRMOFs degrade in the presence of small amounts of water at room temperature [31,111]. In order to find a suitable substitute for linear dicarboxylates, ditetrazolate [112], dipyrazolate [71,113], and di(1H-1,2,3-triazolate) [114] were selected because their acid forms are known to have high pKa values (close to those of carboxylic acids). Likely due to the bridging coordination, MOFs constructed with these linkers have been observed to possess high thermal and chemical stability. Pyrazolate-bridged MOFs have been shown to possess high framework stability even in boiling water, organic solvents, and acidic media [71]. Likely due to the easy mode of synthesis (i.e., [2+3] cycloaddition between azides and organonitriles), tetrazolate-based frameworks have been extensively studied [115]. Compared to other azolates, low basicity (pKa = 4.9 in dimethylsulfoxide) [116] and weak coordination to metals result in relatively low thermal/chemical stability of MOFs constructed with tetrazoles.

19

MOFs with Zr6O4(OH)4 SBU have been of particular interest for potential commercial and industrial applications since they can be easily tailored, and are reported to be chemically and thermally stable. DeCoste et al. [93] have recently shown that there are significant changes in chemical and thermal stability of Zr6O4(OH)4 MOFs with the incorporation of different organic linkers. They observed that as the number of aromatic rings is increased from one to two in 1,4-benzene dicarboxylate (UiO-66, ZrMOF-BDC) and 4, 4 ′ -biphenyl dicarboxylate (UiO-67, ZrMOF-BPDC), the Zr6O4(OH)4 SBU becomes more susceptible to chemical degradation by water and hydrochloric acid. Furthermore, as the linker is replaced with 2,2′ -bipyridine- 5,5′ dicarboxylate (ZrMOF-BIPY), the chemical stability decreases further as the MOF becomes susceptible to chemical breakdown by protic chemicals such as methanol and isopropanol. In contrast to the observations of DeCoste et al. [93], Nickerl et al. [117] have recently shown that the crystal structure of a ZrMOF-BIPY sample soaked in water at room temperature for 24 hours remains essentially identical to that of the untreated sample, as evidenced by the comparison of the PXRD scans of both types of samples. Hence, it can be concluded that structural stability of Zr-MOFs in aqueous media is not only dependent upon the chemical composition of the framework, but also on other factors such as particle size, particle shape, and degree of crystallinity. Volodymyr et al. [118] have employed a strategy that focuses on reducing the ability of the [Zr6O4(OH)4]12+ cluster to coordinate multifunctional linker molecules and, as a beneficial side effect, creates open metal sites in the resultant Zr-containing structure. They successfully utilized a bent dithienothiophene dicarboxylate and a Zr4+ source during a solvothermal reaction to synthesize a Zr-MOF, DUT-51(Zr), containing an eight-connecting Zr cluster with the overall composition [Zr6O6(OH)2(DTTDC)4(BC)2(DMF)6](DMF)12(H2O)19 (DTTDC = dithieno[3,2-b; 20

2′,3′ -d]-thiophene-2,6-dicarboxylate, BC = benzoic acid, DMF = N, N ′ -dimethylformamide).

Each SBU in DUT-51(Zr) was observed to be interconnected by 8 DTTDC molecules forming an 8-connected robust 3-D network with reo topology [118]. The hydrothermal stability of the activated MOF was investigated by soaking a dried sample in water for 12 hours at room temperature. The PXRD scans before and after the soaking treatment showed no noticeable difference, indicating the robustness of the DUT-51(Zr) framework against hydrolysis, as shown in Figure 6 [118]. The structure of MIL-100(Cr), assembled from the Cr3O(CO2)6 cluster and the tritopic linker BTC, has been successfully determined by the combination of simulation and PXRD studies. In its structure, the oxido-centered chromium trimers have been observed to be interconnected by the tritopic carboxylate linkers along the edges to form the so-called “supertetrahedra”, which is built of four chromium trimers as the vertices and four organic linkers as the triangular faces. Shortly after the appearance of MIL-100(Cr), its isostructural series MIL-100(M) (M = Fe3+, Al3+, and V3+) were prepared by replacing the metals in the inorganic SBUs [119–121]. Such MOFs have been observed to show superior hydrothermal stability, and have found wide applications in adsorption/separation (gas, vapor, and liquid), heterogeneous catalysis, and drug delivery. Recently in the MOF field, dipyridyl linkers have been largely used as pillars to knit 2-D layers into 3-D frameworks. For example, a 3-D pillared-layer MOF with a specifically designed dipyridyl pillared linker (2,5-bis-(2-hydroxyethoxy)-1,4-bis(4-pyridyl)benzene), has been observed to exhibit a locking-unlocking system responsible for its characteristic gate-opening type sorption profiles [122]. Bipyridyl linkers with different lengths, such as 4,4′ dipyridylacetylene (DPA) and pyrazine, have been utilized to synthesize cubic topology MOFs 21

via reticular chemistry approach, which show exceptional selectivity, recyclability, and hydrothermal stability for potential usage as membranes in several industrially relevant CO2 separation applications [123]. The ability to synthesize MOFs from a variety of metalloporphyrin struts has created possibilities for the design of a wide range of porphyrinic MOF materials suitable for catalysis, chemical separations, and energy transfer [124–126]. 5,10,15,20-Tetrakis(4-carboxyphenyl) porphyrin (TCPP) is known to be the most frequently used metalloporphyrin linker in MOF synthesis [127–130]. Proper rotation between the porphyrin plane and phenyl rings, together with symmetry diversity of Zr clusters, has been observed to result in 12,8,6-connection modes of Zr clusters. Due to the stability of the Zr cluster and the multifunctionality of porphyrin, these hydrothermally stable MOFs have been applied as biomimetic catalysts [127,129], CO2 fixators [130], and pH sensors [128]. Feng et al. [127] have employed Fe-TCPP as a heme-like ligand, and highly stable Zr6 clusters as nodes for the assembly of water-stable Zr-MOFs. They successfully synthesized a 3-D heme-like MOF, PCN-222(Fe) (PCN = porous coordination network), shown schematically in Figure 7 [127], characterized by one of the largest known average pore diameter of up to 3.7 nm. The PXRD patterns for the as-synthesized MOF were observed to remain unchanged upon immersion in room temperature water, boiling water, as well as 2M, 4M, 8M, and even concentrated HCl solutions for 24 hours indicating exceptional chemical and dimensional stability, as shown in Figure 8 [127]. Most importantly, PCN-222(Fe) was observed to retain its characteristic framework even after treatment with concentrated HCl, which has rarely been observed for other MOF materials. The Zr6 cluster, which has been observed to be one of the most stable building units for MOF synthesis, was proposed to be responsible for the observed exceptional hydrothermal stability of the PCN-222(Fe) framework 22

[127]. It was further suggested that ZrIV, due to its relatively high charge density, polarizes the O atoms of the carboxylate groups to form strong Zr–O bonds, which are characterized by a significantly covalent character [127]. A chelating effect was further proposed to stabilize the four bonds between FeIII and porphyrin, resulting in relatively stronger coordination bonds, and higher framework resistivity towards moisture and acidic solutions. Porous interpenetrated zirconium–organic frameworks (PIZOFs) constitute a family of MOFs characterized by a significantly high degree of flexibility towards the type of substituent coupled with the organic linker during MOF synthesis. This is most probably attributable to the high strength of the Zr–O bond, the large driving force for the formation of the SBUs, and their high degree of connectivity within the framework. A modular-type linker synthesis, facilitated by a high degree of compatibility with a wide variety of substituents, R1, R2, has been investigated to allow for a more flexible selectivity of functional groups as shown in Scheme 1 [131]. The synthesis of the linkers, HO2C[PE-P(R1,R2)-EP]CO2H (1b–8b), was analyzed to make use of Pd/Cu catalyzed C–C cross coupling, which has frequently been observed to result in high yields, and also shows compatibility with a wide variety of functional groups. The coupling partners, utilized in the synthesis scheme, were 2,5-disubstituted 1,4-dihalobenzenes, and alkyl 4ethynylbenzoate. The resulting diesters, 1a–8a, were saponified to obtain the diacids, 1b–8b, as shown in Scheme 1, whilst the variability of the substituents, R1, R2, was achieved through the use of different 2,5-disubstituted 1,4-dihalobenzenes [131]. The second alternative, i.e., the coupling of 2,5-disubstituted 1,4-diethynylbenzene with alkyl 4-iodobenzoate [132], has been considered less appropriate since it involves: (i) a relatively larger number of consecutive steps required for MOF synthesis, (ii) a comparatively larger number of consecutive steps required to vary the moiety used to functionalize the interior of the PIZOFs, and (iii) variation of the 23

substituents of 1,4-diethynylbenzene, during coupling with alkyl 4-iodobenzoate, resulting in different Glaser coupling products [131]. The resulting family of Zr-MOFs, as a by-product of the former coupling scheme, has not only been observed to show high structural stability towards atmospheric moisture, but also a broad variability of substituents at the organic linker, and a high degree of compatibility towards a wide variety of functional groups [131]. 2.6

Framework dimensionality

Apart from the factors mentioned above, the dimensionality of SBU, and its connectivity within the framework also play a vital role in influencing the overall structural stability in aqueous media, both at ambient and at elevated temperatures. It has been observed that although the Zn4O6+ cluster itself is itself stable in a majority of aqueous media [107], the 0-D bonds (all metal SBUs connected only to organic linkers) it forms with BDC are relatively unstable towards exposure to water vapor [52]. For example, MOF-69C which can be synthesized directly from MOF-5 following a very simple post-synthetic procedure, has been observed to undergo complete crystal degradation on exposure to water vapor [52]. It can thus be safely concluded that, in addition to the ligand displacement as discussed earlier, the SBU itself is susceptible towards structural degradation when exposed to an aqueous medium, especially at higher temperatures. This may be due to the alternating pairs of edge-sharing, 4-coordinate Zn cations, and a 6-coordinate, corner-only-sharing Zn cation, connected in the 1-D chains present in the MOF-69C SBU, as shown in Figure 9 [52]. Alternatively, MOF 508-b [108] or Zn-BDCDABCO [102], both of which contain SBUs of two 5-coordinate zinc cations, bridged in a paddlewheel-type configuration such as in HKUST-1, have been evaluated to be only slightly more water-stable than MOF-5. HKUST-1 itself has been observed to be more stable than Zn2+

24

containing MOFs, which is consistent with the general observation that Cu2+ aqueous coordination complexes are more water-stable than the corresponding Zn2+ complexes [109]. Another interesting Zn-containing SBU is the infinite rod present in MOF-74 [110], in which the metal centers exist as edge-sharing octahedra, bridged by the carboxylic acid and hydroxyl functional groups on the dihydroxybenzenedicarboxylic acid (dhBDC) linker, as shown in Figure 10 [52]. In contrast to the weakly assembled chains in MOF-69C as discussed earlier, the chains in MOF-74 are expected to be relatively stronger due to the existence of only edgesharing between the metal centers, as well as coordination to two types of functional groups on each linker. When solvated, all Zn ions in MOF-74 are observed to be 6-coordinate [110], and hence it can be expected that the Zn–O bonds of the linkers will be less susceptible to displacement by incoming water molecules. In the fully de-solvated (activated) state, each Zn ion has also been observed to have a coordinatively unsaturated or open metal site. Upon exposure of this site towards aqueous media, the subsequent removal of water molecules is expected to involve an endothermic reaction with a relatively large energy barrier before actual interaction with a metal-ligand bond. Therefore, water molecules residing in the second coordination shell only, are expected to displace a ligand from the metal center and consequently disrupt the framework. As a consequence, MOF-74 has been predicted and experimentally verified to possess high structural stability on exposure to liquid water [110]. In the early stages of interaction with an aqueous medium, the open metal site allows the water molecules to coordinate without displacing the Zn-linker bonds, thus resulting in the more stable 6-coordinate situation discussed earlier. This suggests that an initial partial exposure, towards a few water molecules, may actually increase the overall stability of the MOF-74 framework towards a linker displacement reaction. This observation has also been found to be in agreement with a recent

25

report on the exposure of HKUST-1 to water vapors at ambient conditions [41]. In general, it is observed that for Zn2+–containing SBUs, the manner in which each metal center is coordinated to other metals as well as to the linker(s) present in the SBU, are both crucially important factors in the context of hydrothermal stability of the framework on exposure to an aqueous medium [52]. 3. Methods to improve hydrostability of MOFs

This section describes the various methods which have been developed and proposed so far for improving the structural stability of MOFs in aqueous media. It is not proposed to include all the possible methods that have been reported to date; however, it still provides a broad overview of the key methodologies which have been adopted recently for the development of hydrostable MOFs for use at both ambient and elevated temperatures. 3.1

In-situ functionalization of organic linker

The functionalization of the organic linker (or ligand) with hydrophobic moieties, during MOF synthesis, has been proposed to enhance the overall stability of the framework in aqueous media. As one of the first attempts of ligand functionalization in MOF structures, Yang et al. [42] have reported the functionalization of the BDC linker of MOF-5 with one or two methyl (CH3) functionalities, during in-situ MOF synthesis via solvothermal route. The resulting methyl- and 2,5-dimethyl-modified MOF-5s were observed to show the same topology as that of MOF-5, supported by the high degree of correspondence between the PXRD patterns of the unmodified and the modified versions of the framework. Both the structures of CH3-MOF-5 and diCH3-MOF-5 were observed to remain stable up to 4 days of exposure to ambient air, as verified by the PXRD patterns of both types of modified MOF-5s shown in Figure 11 [42]. Since 26

2-methylterephthalic acid and 2,5-dimethylterephthalic acid are readily available, this functionalization strategy can be considered viable towards the development of hydrothermally stable MOF-5s on a commercial scale. Trevor et al. [133] synthesized an isoreticular series of porous Nb-O type MOFs with different dialkoxy-substituents of formula Cu2(TPTC-OR) (TPTC-OR = 2′,5′ -di{alkyl}oxy[1, 1′ : 4′,1′′ -terphenyl]- 3,3′′,5,5′′ -tetracarboxylate, R = Me, Et, nPr, nHex). Hydroquinone was utilized as

the starting material to design and synthesize four rectangular planar terphenyl tetracarboxylate organic linker precursors (H4TPTC-OR), by grafting different dialkoxy substitutions on the central phenyl ring, as shown in Scheme 2, with the objective of improving the hydrophobic properties of the framework. Incorporation of these hydrophobic moieties, prior to MOF synthesis, was observed to facilitate a high degree of compatibility of the functionalized linkers with the metal center, without imposing any adverse effects on the structure of the framework [133]. A series of four Nb-O type Cu(II)-based MOFs were synthesized with different alkoxy substitutions, which were observed to enhance both the moisture and thermal stability of the isoreticular framework, as well as its H2 adsorption capacity [133]. The role of pendant alkoxy groups, in evaluating the moisture stability of the four Cu2(TPTC-OR) MOFs, was evaluated through PXRD studies, and the nature of interaction of water droplets with the functionalized frameworks. It was observed from the PXRD data that the –OMe substituted framework is the least stable of the studied systems, and begins to decompose during the activation process, indicated by the increasing intensity of the reflection at 2θ ~ 36.4o, shown by the PXRD scan in Figure 12 [133]. This reflection was also observed in the as-synthesized pattern, and was proposed to be attributable to the initial stages of framework decomposition (formation of CuO). The other three structures were, however, observed to retain crystallinity upon solvent removal, 27

activation, and a minimum of 1 hour exposure to atmospheric conditions (50% RH) after activation. The deviation of the reflection intensities of the as-synthesized forms from the simulated pattern, led to the conclusion that the single crystal data include no contributions from solvent molecules residing inside the pores, while deviations apparent in the activated structure were proposed to arise from slight distortions in the framework to minimize the energy at the end of the activation process [133]. The PXRD pattern for the –OEt substituted framework, obtained after 16 days of exposure to atmospheric conditions (50% RH), showed significant broadening of diffraction peaks coupled with a decrease in the signal-to-noise ratio, indicative of considerable framework distortions at the initial stages of decomposition [133]. In contrast, the minimal broadening of reflection peaks, observed in the PXRD pattern of the –OnHex substituted framework, led to the conclusion that it is relatively more stable towards atmospheric moisture [133]. The stability of the –OnPr substituted framework was observed to fall between those of the –OEt and –OnHex substituted frameworks, as suggested by the relative broadening of diffraction peaks; however, an additional reflection indicative of the formation of CuO was observed in its PXRD pattern – a feature not observed in the expectedly less stable –OEt substituted framework. The relative moisture stability was further investigated via the interaction of water droplets on the surfaces of each of the three types of frameworks. When a single droplet of water was dropped on the surface of a ~10 mg powdered MOF, both the –OEt and –OnPr substituted frameworks were observed to absorb the water droplet as shown in Figure 13 [133]. However, in case of the –OnHex substituted MOF, the surface of the water droplet was observed to become fully coated with the powder, retaining its shape completely before falling off the surface of the powder, and thus indicating an exceptionally hydrophobic (superhydrophobic) material, as shown in Figure 13 [133].

28

Wu et al. [39] synthesized a series of MOFs with a cubic topology and the ligand functionalized with water-repellant groups R1 and R2 (R1 = trifluoromethoxy, R2 = H). The resultant MOF, Banasorb-22, was tested for its hydrothermal stability by exposing it to steam over boiling water for one week. The PXRD patterns of the MOF samples, before and after exposure, showed no change in peak positions as a result of the steam exposure but changes in peak intensities at 6.78o and 9.6o, as shown in Figure 14 [39]. It was observed from BET measurements that the surface area of Banasorb-22 reduced from 1113 m2g-1 to 210 m2g-1, after 1 week exposure to steam. In contrast, under the same conditions, the BET surface area of the nonfunctionalized MOF decreased from 2365 m2g-1 to 50 m2g-1 in just a few minutes, confirming that the water-repellant nature of R1 and R2 successfully contributed towards the increased hydrophobicity of the framework [39]. In an effort to construct functional porous MOFs, Zhang et al. [134] recently reported that the reaction of Ni(II) anions with 2,4,6-tri(4-pyridinyl)-1,3,5-triazine (TPT) and o-phthalic acid (OPA) as co-ligand, can result in the production of MOFs with a high degree of porosity. Furthermore, the decoration of the pore surface and tuning of adsorption properties could be achieved by introducing different functional groups on the OPA ligand. Based on this strategy, they synthesized a new class of MOFs, TKL-101 to 107 (TKL = Tianjin Key Lab of Metal and Molecule-based Materials), with different functionalized (–NH2, –NO2, and –F) OPA as coligands, as shown in Figure 15 [134]. PXRD patterns, obtained after the degas treatment, revealed that TKL 101–104 failed to retain their long-range order, while TKL 105–107 still presented high crystallinity after adsorption experiments, indicating good stability of the fluorine-decorated frameworks after the removal of guest solvent molecules. Furthermore, on comparing the PXRD patterns of TKL-104 with those of TKL 105–107, it was observed that for 29

achieving higher hydrothermal stability of the framework, functionalization with F-atom at the 3position of the OPA ligand is more effective than at the 4-position. Taylor et al. [135] reported the synthesis of a hydrophobic phosphonate monoester MOF, barium tetraethyl-1,3,6,8-pyrenetetraphosphonate, BaH2L (CALF-25, CALF = Calgary Framework), which was proposed to derive water vapor stability from the protective ethyl groups lining the pores. To assess the hydrophobic nature of CALF-25, water vapor sorption experiment was performed at 5 K intervals, between 298 and 313 K, as shown in Figure 16 [135]. The water adsorption isotherms were observed to show type-III behavior, indicative of a relatively lower level of affinity between the surface and the adsorbate molecules. The heat of adsorption was calculated to be ~45 kJ mol-1 across all loadings of water; however, a zero-loading value was not computed owing to insufficient availability of low pressure data [135]. The observed very low value for the heat of adsorption of water suggested that the pore surface is essentially hydrophobic. Since CALF-25 has been evaluated to be an acidic material composed of metal cations and oxo anions, such a low value for the heat of adsorption of water is indicative of the fact that the ester groups lining the pore surfaces effectively shield the polar, acidic barium phosphonate chains behind the less polar ethyl groups, thus making the pore surface hydrophobic [135]. Exposed metal cations within MOFs have been demonstrated to lead to outstanding properties for hydrogen storage [136], gas separations [137–142], and catalysis [12]. Among the azolate-based MOFs of this type, Mn3[(Mn4Cl)3(BTT)8]2.20MeOH (Mn-BTT, H3-BTT = 1,3,5tris(2H-tetrazol-5-yl)benzene), a rigid, high-surface area MOF with exposed Mn2+ sites, has been investigated to exhibit a high Lewis acid catalysis [143]. However, the relatively low waterstability of this tetrazolate-bridged framework has constrained its use in applications which also 30

require higher hydrothermal stability. In an effort to enhance the hydrothermal stability of tetrazolate-bridged frameworks, Valentina et al. [70] proposed their pyrazolate-bridged analogues, and reported that reactions between the tritopic pyrazole-based ligand, 1,3,5-tris(1Hpyrazol-4-yl)benzene (H3BTP), and transition metal acetate salts in DMF, result in microporous pyrazolate-bridged MOFs of the type M3(BTP)2.xsolvent (M = Ni, Cu, Zn, and Co), which adopt the Mn-BTT structure and feature exposed metal cation sites. They observed that the assynthesized Ni3(BTP)2 framework is stable to all of tested environments (water, acids, or base for two weeks at 100oC), and completely retains both its crystallinity, surface area, and porous nature after 14 days of uninterrupted test reactions [70]. PXRD data, collected before and after each test, were used to confirm its structural chemical integrity as shown in Figure 17 [70]. Other stability studies on tetrazolate- [112], triazolate- [141], and pyrazolate-based [71,142] frameworks have been performed, but the chemical stability of each in both acidic and basic media was either observed to be either inferior to Ni3(BTP)2 or not reported [70]. In contrast, the Cu3(BTP)2 and Zn3(BTP)2 frameworks were observed to undergo transformations to non-porous crystalline solids upon extreme chemical treatment, as generally expected of MOFs with Cu and Zn metal centers. As depicted in Figure 18 [70], Cu3(BTP)2 was observed to show a progressive phase transition in boiling water, converting to Cu3(BTP)2.6H2O upon refluxing in aqueous NaOH (pH = 14), or HCl (pH = 3) solutions for one day. The highest resistance of the Cu3(BTP)2 framework to pH 14 was observed to be one day, and it was further found to be stable for two weeks in benzene [70]. Despite its extremely high thermal stability, the Zn3(BTP)2 framework was observed to show resistance towards hot acidic media, but not to the extent demonstrated by Ni3(BTP)2. Although it was observed to remain stable upon heating at 100oC in pH 3 aqueous HCl for seven days as shown in Figure 19 [70], the structure of Zn3(BTP)2 showed gradual

31

structural degradation upon a similar treatment at pH 2. In addition, it was observed to react in water and especially in basic solutions, transforming into the cubic phase Zn12 [Zn2(H2O)2]6(BTP)16 [70]. Lanthanide-based MOFs (Ln-MOFs) have been reported to be the first example of exceptionally water-stable multifunctional materials [144]. Zhou et al. [145] synthesized two novel isostructural Ln-MOFs – [Eu2(BPDC)(BDC)2(H2O)2]n and [Tb2(BPDC)(BDC)2(H2O)2]n via a mixed-ligand approach using 2,2′ -bipyridine- 3,3′ -dicarboxylic acid (H2BPDC) and 1,4benzenedicarboxylic acid (H2BDC) under hydrothermal conditions. The structural stability of the as-synthesized MOFs was investigated by immersing each separately in 0.01M aqueous solutions of sodium chloride, sodium fluoride, sodium bromide, and sodium iodide for 2 days. Figure 20 shows the PXRD patterns of both kinds of MOFs in the as-synthesized states, and after immersion in each of the four aqueous solutions for 2 days [146]. As evident from the figure, no noticeable differences can be observed from the PXRD patterns between the structures of the assynthesized MOFs, and of the ones recovered after immersion in each of the four aqueous sodium halide solutions for 2 days. Liu et al. [147] constructed a 12-connected neodymium trimesate [Nd(trimesate)] MOF by linking the Lewis acid sites of tetranuclear neodymium carbonate cluster and organoboronderived tricarboxylate bridging ligand. The [Nd(trimesate)] MOF was synthesized and activated according to the published literature [148]. Investigation of the chemical stability of the activated MOF was performed by heating the sample in boiling benzene, boiling methanol, and boiling water for 1 week (conditions that reflect potential extreme industrial requirements). As evident in Figure 21 [149], no detectable changes could be observed in the PXRD pattern of the activated MOF after treatments in each of the three solvents under the designated conditions, indicative of

32

a highly moisture-resistant and chemically robust framework structure. The BET surface area of the MOF sample which was treated in boiling water for 1 week was observed to be 403 m2g-1 compared to that of the activated sample of 585 m2g-1, further validating the framework stability and robustness of Ln-MOFs in aqueous media under extreme conditions [147]. 3.2

Post-synthetic modification of organic linker and/or metal center

Once synthesized, the physical and chemical properties of MOFs can be further tuned by the incorporation of functional groups on the organic linker, or on the unsaturated metal sites in the framework through PSM [150–151]. PSM can thus be broadly defined as chemical derivatization of MOFs after their syntheses, which in a majority of cases has been achieved via covalent bond formation with the framework [152]. The potential advantages offered by the PSM approach for the development of water-stable MOFs can be appreciated by the following considerations [152]: (i) it is possible to include a more diverse range of water-repellant functional groups, freed of the restrictions posed by MOF synthetic conditions, (ii) a given MOF structure can be modified with different reagents, thereby generating a large number of topologically identical, but functionally diverse water-stable MOFs, and (iii) control over both the type of substituent and the degree of modification allows introduction of multiple water repellant functionalities into a single framework in a combinatorial fashion, enabling an effective way to maximize the hydrothermal stability of the structure. Amine groups are very versatile, and have been introduced into many MOF structures through the use of 2-amino-1,4-benzenedicarboxylate (BDC-NH2) as linker in MOF synthesis. The amine group on the linker not only helps to improve affinity for particular gases such as CO2, but also provides a platform for developing further functionalized MOFs via PSM. A wide range of functionalities have been incorporated into MOFs following reaction with an amine 33

group on the pore wall [150]. It has been observed that amine-containing MOFs can readily undergo PSM to form amide-functionalized MOFs [153–154]. It was hypothesized that the introduction of hydrophobic alkyl chains via PSM is expected to result in an improved moisture resistance, and relatively higher hydrophobicity of these frameworks. IRMOFs are cubic frameworks, comprised of Zn4O clusters and dicarboxylate ligands. Nguyen and Cohen [43] successfully synthesized IRMOFs using 1,4-benzenedicarboxylate (IRMOF-1), and 2-amino-1,4benzenedicarboxylate (IRMOF-3), and investigated the PSM derivatives of the latter. The second MOF system examined belonged to the MIL class of materials [24], specifically MIL-53(Al)NH2, a flexible MOF comprised of infinite Al3+ clusters and NH2-BDC linker. The IRMOFs have generally been observed to be sensitive towards moisture, while the MILs are generally known to be stable towards water and other polar solvents. In order to assess changes in the moisture-stability and hydrophobicity/philicity of the framework following PSM, each material was exposed to ambient air, or immersed in water, and then characterized using contact angle measurements, PXRD, and Scanning Electron Microscopy (SEM) [43]. From contact angle measurements, IRMOF-1 and IRMOF-3 were observed to be hydrophilic, indicated by a high degree of water absorption and water contact angles close to 0o. Similarly, IRMOF-3-AM1, AM2, and -AM3 were also observed to display contact angles close to 0o (AM1 = ‘amide’ with one-carbon chain). However, IRMOFs modified with longer alkyl substituents (e.g. IRMOF-3AM4 and longer), were observed to show contact angles ≥ 116o, clearly indicating the hydrophobic nature of these materials. The contact angles of these hydrophobic samples were observed to remain unchanged even after exposure to ambient air for several weeks [43]. MOFs with smaller alkyl chains were observed to require a higher degree of functionalization, in order to effectively introduce hydrophobicity within the framework. For example, a partially

34

functionalized IRMOF-3-AM15, with only 25% of the amine groups converted to amides, was observed to display a high degree of hydrophobicity as suggested by contact angle measurements [43]. In contrast, for the IRMOF-3-AM4 framework to be hydrophobic, atleast 50% of the amine sites needed to be converted to amide groups. Branched alkyl substituents were also observed to enhance framework hydrophobicity, as IRMOF-3-AM3 was observed to be hydrophilic while IRMOF-3-AMiPr was observed to be hydrophobic. It was deduced from these experiments that both the chemistry and the structure of the functional group, as well as the degree of functionalization, bear crucial significance in improving the hydrophobic properties within MOF materials [43]. Inspite of the high hydrothermal stability of the UiO-66 MOF family in acidic conditions, their stability in basic media has been found to be less impressive. The difference in the stability with respect to the substituent group on the terephthalic acid linker has already been reported [91,155]. According to these studies, UiO-66-NO2 and UiO-66-(CH3)2 show high stabilities in both strongly acidic and basic media, UiO-66 and UiO-66-Br show less crystallinity after soaking in NaOH (pH = 14), whilst UiO-66-NH2 gives rise to a complete MOF decomposition under the same basic conditions. In order to increase the stability of the UiO-66-NH2 framework in basic media, Jordi and Darren [156] recently reported a facile strategy for its PSM by the incorporation of a resonant indole moiety into the structure. They proposed that diazotization of the amino groups, to obtain UiO-66-N=N-ind, could potentially improve the UiO-66-NH2 stability in basic media. The PSM of UiO-66-NH2 was carried out following a diazotization reaction, using 1-methylindole as a coupling agent, to form UiO-66-N=N-ind [156]. The BDCNH2 linkers were easily diazotized with NaNO2 and HCl, and then reacted with the indol compound for 1 hour to obtain UiO-66-N=N-ind1h, in order to avoid the azo-coupling attack of 35

the diazonium ion on the nitrogen of the indole (to form diazoaminobenzene-like structures). The UiO-66-N=N-ind1h framework was observed to show high stability in acidic conditions, and an increased chemical stability at basic pH, as suggested by the PXRD patterns shown in Figure 22(a), and N2 adsorption isotherms in Figure 22(b) [156], which clearly indicate that higher degrees of crystallinity and porosity over UiO-66-NH2 were retained up to pH 11. This enhancement in structural stability at basic pH was proposed to be due to the high resonance of the resulting aryl azocompound (extended delocalized system), which conferred improved chemical tolerance to the framework [156]. Ji et al. [157] recently synthesized a titanium MOF using PSM of the BDC linker, used in the synthesis of MIL-125, with methyl groups. The resultant methyl-functionalized MIL-125 was observed to be isostructural with unmodified MIL-125 as evidenced by PXRD studies of each of the two MOFs. The resistance of both types of MOF samples towards aqueous media was measured by immersion in water for 1, 2, and 4 hours at room temperature. Table 4 presents the changes in the surface areas and pore volumes of both the unmodified and methyl-modified MOF samples after immersion in water for the designated time periods [157]. As shown in the table, the BET surface area of unmodified MIL-125 decreased by 167%, while the total pore volume reduced by 85% after immersion in water for 4 hours. In contrast, the BET surface area of methyl-modified MIL-125 was observed to decrease by 52%, and the total pore volume by only 2.5%, after measurements were taken on samples recovered after immersion in water for 4 hours. As shown by the PXRD scans in Figure 23(a), the diffraction intensity of the original peaks of unmodified MIL-125 quickly and significantly decreased after water immersion for 1 hour, while the recovered sample was observed to be completely amorphous after immersion for 4 hours, indicating complete collapse of the framework structure [157]. In comparison, the 36

sample of the methyl-modified MIL-125 resulted in virtually unchanged diffraction peaks after being immersed in water for 1 hour, and even the sample after immersion for 4 hours yielded sufficiently strong peaks for correct identification of the characteristic MIL-125 framework, as shown in Figure 23(b). These results clearly demonstrate that the introduction of methyl groups to a titatnium MOF via PSM approach improves the structural stability against water, as has also been observed before for methyl-functionalized MOF-5 [42]. 3.3

Ligand Pillaring

Pillared MOFs have recently been widely explored as a type of platform for gas storage/separation, and are found to consist of 2-D layers with ligating sites for linkage with ditopic pillars, usually of dipyridine derivatives, into 3-D architectures [102,158–160]. The structures and properties of pillard MOFs have been observed to be readily tuned by functionalizing either the linkers within the 2-D layers, or the ditopic pillared ligands. Pillared MOFs are solvothermally synthesized by mixing two types of organic ligands, usually aromatic dicarboxylates and diamines, with a metal salt [73]. Metal ions and dicarboxylates are observed to form 2-D sheets, while diamines act as pillars connecting the sheets in the third dimension [74]. MOFs synthesized in this fashion have produced several examples of non-catenated (e.g. DMOF [102]), as well as catenated structures (e.g. MOF-508 [108]). Lu

et

al.

[161]

synthesized

a

water-stable

pillar-layered

porous

MOF,

{[Cu2(TCMBT)(bpp)(µ3-OH)].6H2O}n (bpp = 1,3-bis(4-pyridyl)propane) from a flexible ligand, N, N ′, N ′′ -tris(carboxymethyl)-1,3,5-benzenetricarboxyamide (TCMBT), using an evolutionary approach in terms of both structure and properties. They were able to successfully control the evolution steps by gradually increasing the complexity of both the organic and the inorganic

37

building blocks, and successfully synthesized four MOF complexes, {[Cu(TCMBT)(H2O)3]2[Cu(H2O)6].3H2O}n with simple layers, {[Cu3(TCMBT)2(H2O)3].2H2O}n with advanced layers, {[Cu3(TCMBT)(py)2(µ3-OH)].2H2O}n (py = pyridine) with augment layers, and porous MOF, {[Cu2(TCMBT)(bpp)(µ3-OH)].6H2O}n with a pillared-layer structure, as shown in Figure 24 [161]. The water stability of the MOF complex synthesized with simple layers was investigated in both room temperature water and boiling water, with an exposure period of two months each. The complex was observed to retain its structure as a result of both of these treatments, as suggested by the PXRD patterns shown in Figure 25 [161]. It was further observed that the MOF sample, after activation, shows no structural degradation after dispersion in water for two consecutive months. This excellent hydrothermal stability was attributed to the presence of a four-member Cu cluster being partially and strongly coordinated to N atoms in the coligands (bpp), instead of being purely coordinated to the O atoms of the carboxylate, which has been observed to be relatively more sensitive on interaction with water molecules [61,71,141]. Liu et al. [162] have recently synthesized two new 3-D frameworks [Zn3(bpdc)3( 2,2′ dmbpy)].DMFx-(H2O)y, and [Zn3(bpdc)3(3, 3′ -dmbpy)].DMF4-(H2O)0.5, by modifying their parent structure [Zn3(bpdc)3(bpy)].DMF4-(H2O) (bpdc = biphenyl-4, 4 ′ dicarboxylic acid; z, z ′ dmbpy = z, z ′ -dimethyl- 4,4′ -bypyridine; bpy = 4,4′ -bypyridine) with methyl-functionalized pillar ligands. In order to investigate the effect of –CH3 groups on the water stability of the assynthesized MOFs, both the as-synthesized frameworks were tested in a water-vapor environment at 323 K. Figure 26 [162] shows that the framework of [Zn3(bpdc)3( 2,2′ dmbpy)].DMFx-(H2O)y was observed to remain unchanged after exposure to water vapor for up to ten days, whereas its parent structure (without –CH3 functionalization) showed significant degradation after exposure over the same length of time. These results show that functionalizing 38

the framework with hydrophobic groups, such as a –CH3, might effectively enhance the hydrothermal stability of certain MOF structures. 3.4

Catenation

Catenation is defined as the interpenetration or interweaving of two or more identical and independent frameworks. Catenated MOFs are observed to be more thermodynamically stable [157,164], and also lead to lower water adsorption loadings [164]. Figure 27 [73] shows a schematic of non-catenated pillared MOFs (e.g. Zn-TMBDC-BPY [165] synthesized using TMBDC and BPY), and two-fold catenation in pillared layer frameworks (e.g. MOF-508 [108] synthesized using BDC and BPY). In the context of MOF synthesis, catenation employs the use of metal-organic building blocks that can be assembled on a triply periodic P-minimal geometric surface, resulting in an interpenetrated or interwoven network which is characterized by a high degree of porosity and hydrothermal stability. Banglin et al. [166] have employed catenation for the synthesis of MOF-14, by using the binuclear copper carboxylate “paddle wheel” [167] as a square SBU and the ortho, meta, and para positions of the benzene ring as a triangular SBU. The catenated framework of MOF-14 was observed to occupy only a small fraction of the available space in the crystal containing large cavities, with the size of each cavity being substantially larger than that of the noninterpenetrated MOF-5 [168]. The separating surface in the MOF structure was observed to be the P-minimal surface, one of the simplest triply periodic minimal surfaces in Euclidean space, which partitions into two interpenetrating labyrinths, each consisting of two orthogonal channels meeting at the sites of a primitive cubic lattice, as shown in Figure 28 [166]. Guest exchange studies on this MOF have indicated that it is structurally stable in moist air and highly insoluble in water [166]. 39

Jasuja and Walton [73] have recently demonstrated that, by using catenation in combination with a pillaring strategy, it is possible to obtain water-stable MOFs even when the pillar ligand is characterized by a lower basicity. They showed that after 90% RH exposure, a comparison of Zn-BDC-DABCO (DMOF) and Zn-BDC-BPY (MOF-508) revealed that MOF508 is relatively more stable due to its two-fold interpenetration that prevents significant water adsorption. In contrast to the known instability of DMOF [75], the PXRD patterns of activated, water-exposed, and regenerated (activated after water exposure) samples of MOF-508 were observed to remain unchanged, as shown in Figure 29 [73]. As reported in Table 5, DMOF was observed to undergo a complete loss in its surface area upon exposure to 90% RH, while MOF508 did not [73]. In this work, it was observed that MOF-508 shows a higher degree of waterstability (after 90% RH exposure), in spite of being synthesized with a pillar ligand (BPY) of lesser basicity (pKa = 4.16 [169]), compared to DABCO (pKa = 8.86 [169]) which is used in DMOF. As mentioned before, the Lewis acid metal sites are expected to form stronger bonds with more basic ligands, supporting the hypothesis that DMOF should be more water stable than MOF-508. However, MOF-508 has been experimentally observed to be more stable than DMOF due to the existence of a two-fold catenated framework [73]. The same logic has also been found applicable to the ambiguous water-repellant nature of the SNU-80 framework, synthesized by Xie et al. [164]. 3.5

Plasma enhanced chemical vapor deposition of perfluroalkanes

Plasma enhanced chemical vapor deposition (PECVD) of perfluoroalkanes is one of the most recent techniques to modify the wetting properties of surfaces for desired applications. However, in case of materials characterized by extraordinary high surface areas, such as MOFs, application of PECVD treatment has been practically experienced to be slightly more 40

challenging. The purpose of the PECVD treatment is to attach fluorinated groups on the surface of the MOF with the purpose of increasing the overall hydrophobicity of the framework [84]. The foremost advantage of this technique is that it can be applied to any MOF-containing aromatic hydrogens, and can be used to attach a variety of functional groups on the inner surfaces of MOF structures. However, practical implementation of this technique is more difficult than conventional functionalization methods due to the formation of highly reactive radicals in the plasma which not only react with the aromatic hydrogen atoms, but also with the CFx groups already functionalized onto the MOF surfaces. Decoste and Peterson [171] have presented a treatment of the as-synthesized Cu-BTC (HKUST-1) with PECVD of perfluoroalkanes, with the aim of improving the hydrothermal stability of the parent framework. In order to assess the hydrophobicity of the modified MOF, a 0.5 g powder sample of Cu-BTC, treated with hexafluoroethane (C2F6) for 4 hours at a pressure of 0.3 mbar and a plasma power of 50 W, was placed on top of deionized water to examine its floatability, and also pressed in the form of a pellet to measure its surface wettability. The results, shown in Figure 30, clearly indicate the extent of hydrophobicity of the treated Cu-BTC, which was observed to float completely on the surface of water and resulted in a water contact angle of 123o, as opposed to the untreated sample which dissolved inside water and yielded a contact angle of 59o [171]. Hence it can be concluded that the presence of CFx groups, on the surfaces of the pores, adds to the hydrophobicity of the material causing it to repel water molecules [171]. In order to investigate the hydrothermal stability of the treated MOF under harsh humidity conditions, both Cu-BTC and C2F6 plasma-treated Cu-BTC samples were rapidly aged at 45oC and 100% RH for three days. The PXRD patterns, obtained following the aging treatment, revealed a near-complete change in the structure of the untreated sample, as shown in 41

Figure 31 [171]. The plasma-treated sample, however, showed a negligible change in the structure, confirming the highly hydrophobic nature of the fluorinated groups functionalized onto the surface of the MOF and an improved overall stability of the framework. Decoste et al. [84] also treated Cu-BTC with a PECVD of perfluorohexane with the aim of creating a hydrophobic form of the framework. Upon immersion in water at room temperature, the PXRD pattern of CuBTC was observed to change completely over a period of 24 hours, while the PXRD pattern of the plasma-treated sample virtually remained unchanged as shown in Figure 32 [84]. 3.6

Thermal treatment

Among the MOFs reported to date, Zn-based MOFs, particularly IRMOF-1, have been observed to be most moisture-sensitive owing to their relatively weak metal-oxygen coordination bonds, which are easily prone towards hydrolysis, and to finally lead to disintegration of the framework in aqueous media [38]. Yang and Park [172] reported, for the first time, that carboncoating the IRMOF-1 surface leads to a considerable enhancement in the moisture resistance of the entire framework. In previous studies, “carbonaceous grease”, which is known to be more hydrothermally stable than MOFs, has been incorporated into such frameworks to help stabilize them against hydrolysis [173–174]. Recently, a variety of MOFs have been widely exploited as templates, or precursors for nanoporous carbonaceous materials [175–177]. In the intermediate state of the transformation, a simple heat treatment at a specific temperature was observed to cover the IRMOF-1 framework in an amorphous carbon coating, which shielded the direct exposure of the framework to moisture, thereby preventing hydrolysis as shown in Figure 33 [172]. Higher temperatures were experienced to produce more moisture-resistant MOFs due to the possibility of thicker coatings; however, overheating was observed to result in a ZnO@carbon material which failed to display microporosity, and the intrinsic water adsorption 42

characteristics of MOFs [172]. The effects of carbon-layer coating on the stability of the IRMOF-1 framework, in terms of resistance to hydrolysis, were predicted by immersing both the untreated and the thermally treated samples in water, as shown in Figure 34 [172]. The wellfacted cubic crystals of IRMOF-1 were observed to hydrolyze rapidly to form powders within 10 s, verified by the high turbidity of the initially transparent water, thus indicating the high moisture sensitivity of the untreated framework. In contrast, the water above the thermally treated sample was observed to remain colorless for the entire duration of the 2 hour soaking time [172]. The de-solvated MOFs were exposed to ambient air (34% RH), and subsequently evaluated using PXRD to assess the degree of retention in crystal structure in the presence of moisture [43,84,173]. The PXRD patterns, shown in Figure 35, verify that the intrinsic crystal structure of thermally modified IRMOF-1 was preserved even after a two-week exposure to atmospheric moisture, indicating the improved hydrophobicity of the carbon-coated framework [172]. Exposure of the de-solvated IRMOF-1 to air, for three days, was observed to result in a shift in the relative peak intensities, as well as the appearance of new peaks, indicating commencement of hydrolysis of the framework [194]. Further exposure was observed to result in the near-complete disappearance of the original patterns, indicative of a virtually complete phase transformation. In contrast, the patterns of all the thermally-modified MOFs were observed to remain essentially unchanged even after 14 days of exposure, with appearance of no new peaks and no reduction in peak intensities [194]. Thus, the amorphous carbon coating, introduced via thermal modification, resulted in an increased resistance of the moisture-sensitive IRMOF-1 framework towards atmospheric moisture.

43

3.7

Metal ion doping

Incorporation of metal ions into MOFs has also been considered one of the ways to enhance the water-stability of the framework. Since metals with relatively higher oxidation states are expected to be more water-stable as discussed earlier, doping of such metal ions into a moisture-sensitive framework is also expected to enhance its overall resistance towards hydrolysis. Li et al. [178] synthesized two chemical compositions of Ni-doped MOF-5s using a solvothermal crystallization process – Zn3.48Ni0.52O(BDC)3(DMF)2.6 (Ni13-MOF-5) and Zn3.52Ni0.88O(BDC)3(DMF)2.2 (Ni22-MOF-5). As indicated by the PXRD patterns shown in Figure 36 [178], the sample of undoped MOF-5 was observed to show a decreasing trend in crystallinity on exposure to ambient air (30–37% RH), as the time of exposure was increased gradually from 20 minutes to one week. Moreover, an extra peak at 2θ = 8.82o was also observed in the PXRD pattern of undoped MOF-5 after 2- day exposure, indicating the initial stages of decomposition of the framework, as shown in Figure 36(b). After one week of exposure to ambient moisture, no diffraction peaks corresponding to the intrinsic MOF-5 structure could be observed, suggesting complete structural transformation of the framework to ZnBDC.xH2O [31,40]. In contrast, the decomposition of Ni13-MOF-5 framework was observed to be much slower compared to that of undoped MOF-5, as suggested by the appearance of the diffraction peak at 2θ = 8.82o, which was observed after a four-day exposure to air, as shown in Figure 36(c) [178]. Furthermore, diffraction peaks corresponding to the intrinsic MOF-5 structure were still clearly identifiable in the PXRD pattern of Ni13-MOF-5, after one week of exposure. In case of Ni22-MOF-5, the diffraction peak at 2θ = 8.82o was not observed at all after one week of

exposure to moist air, indicating even higher framework stability than that observed for Ni13MOF-5 [178].

44

3.8

In-situ synthesis of MOF nanocomposites

Another way of improving the hydrostability of MOFs is to either incorporate them in other hydrophobic matrices, or the incorporation of hydrophobic second-phase nanoparticles within the framework during MOF synthesis. However, the method of in-situ incorporation has to be designed in such a way that the second-phase nanoparticles are uniformly implanted into the matrix, and least interfere with the intrinsic porosity of the framework in order to preserve the characteristic gas/water adsorption/desorption behavior of the structure. The use of carbon nanotubes (CNTs), as second-phase nanoparticles, has been employed in numerous applications for synthesizing nanocomposites with improved electroconductive, thermoconductive, mechanical, and hydrophobic properties [179]. Yang et al. [174] incorporated pre-functionalized multiwalled carbon nanotubes (MWCNTs) into MOF-5, using an in-situ synthesis procedure, to synthesize a hybrid nanocomposite (denoted as MOFMC). The PXRD patterns of neat MOF-5 and MOFMC were observed to be in perfect agreement with prior literature on the MOF-5 lattice [180], confirming that the incorporated MWCNTs did not disturb the intrinsic crystal structure of the framework [195]. Moisture stability tests on MOF-5 and MOFMC were carried out at 33% RH and 23oC, as shown in Figure 37 [174]. Exposure of de-solvated MOF-5 to air, for 2 hours, was observed to result in the appearance of a new XRD peak at 2θ ≈ 8.4 o , indicating the commencement of structural decomposition of the framework [31,38,53]. The intensity of the new peak was observed to increase gradually with the increasing time of exposure, indicating acceleration in the rate of decomposition of the MOF-5 framework. Following one week of exposure to ambient air, the decomposition of the framework was observed to be almost complete, as the PXRD scan did not include any reflections characteristic of the intrinsic MOF-5 crystal structure. In contrast, the reflections corresponding to MOFMC were not observed to 45

change even after one week of exposure to moist air, clearly indicating the enhancement in moisture-stability of the MOF-5 framework due to the incorporation of MWCNTs [174]. In an effort to eliminate the pre-functionalization step of the second-phase nanoparticles or the matrix for the in-situ fabrication of hydrophobic MOF nanocomposites, Wu et al. [181] recently presented a new synthetic strategy for the in-situ fabrication of MOF-5@SBA-15 nanocomposite. The fabrication approach relied on the direct physical dispersion of MOF-5 precursors into the pores of SBA-15 using the solvothermal method [182–183], thus eliminating the complicated pre-functionalization of the substrates employed in previous studies [174,184– 186]. The proposed synthetic scheme of MOF-5@SBA-15 nanocomposite is shown in Figure 38 [181]. The moisture stability tests, for the as-synthesized MOF-5@SBA-15 nanocomposite, were carried out at ~50% RH and 25oC. The PXRD scans of MOF-5 and MOF-5@SBA-15 samples, as a function of exposure time, are shown in Figure 39 [181]. After exposure to ambient air for 36 hours, an additional peak in the PXRD pattern of MOF-5 was observed at 2θ ≈ 8.4 o , corresponding to the decomposition of de-solvated MOF-5 crystals [31], as shown in Figure 39(a) [181]. Consequently, the intensity of the peak at 8.4o was observed to increase with increasing time of exposure which, along with the decrease in intensity of the peaks at 6.8o and 9.6o, indicated acceleration in decomposition of the MOF-5 framework. After one week of exposure, the peaks attributed to MOF-5 crystal were observed to disappear completely, indicating complete disintegration of the structure under the selected atmospheric conditions. In contrast, the PXRD pattern of MOF-5@SBA-15 was observed to remain unchanged even after one week of exposure to ambient air, as shown in Figure 39(b). These results imply that the robust, porous structure of SBA-15 enhanced the stability of the otherwise moisture-sensitive MOF-5 crystals by confining them within the pores of the SBA-15 matrix [181]. Hence, 46

encapsulation of moisture-sensitive MOFs within pores of more hydrophobic matrices might also be considered an efficient strategy for improving the overall hydrostability of the structure, since it not only imparts hydrophobicity to the framework, but also eliminates the pre-functionalization steps for both the encapsulant and the host matrix. 4. Hydrothermal Cyclic Stability

In certain applications where MOF-based adsorbents are subjected to continuous alternating phases of adsorption and desorption, like sorption-based chillers or heat pumps, the selected framework is not only required to be hydrostable, but also capable of retaining its structural integrity on repeated exposure to subsequent adsorption/desorption cycles [27,187– 190]. Hydrothermal cyclic stability of a MOF-based sorption material ensures that it not only retains its structural stability during the hydrothermal cycling process, but also maintains the adsorption capacity of the refrigerant selected for the chiller or the heat-pump application, for which it has been selected as an adsorbent. Hence, the requirement of hydrothermal cyclic stability poses additional constraints on the method being adopted to enhance the hydrostability of the framework, so as to ensure sufficient level of metal-ligand bond strength for achieving satisfactory performance in such applications. A number of research attempts have recently been focused on the PSM of MOFs for their intended use as potential adsorbents in sorption-chiller or heat transformation applications [190–193]. Felix et al. [191] synthesized two amine-modified MOFs, H2N-UiO-66-Zr and H2N-MIL125-Ti, and investigated their hydrothermal cyclic stabilities for intended use as potential adsorbents in heat transformation applications. The multi-cycle ad-/desorption experiments were performed as follows: a humidified argon gas flow (5.6 kPa H2O vapor pressure) was conducted through a thermogravimetric balance. The temperature of the sample chamber was varied 47

between 40oC and 140oC, with a cycle time of 5 hours. The initial or average water loading lift under the cycling conditions was observed to be 0.45 gg-1 for H2N-UiO-66-Zr, and 0.4 gg-1 for H2N-UiO-66-Zr. However, it was observed that the thermodynamic water sorption equilibrium was not reached within the individual cycles. Consequently, a 20 hour thermodynamic analysis was interposed before the cycling experiment, after 20 and 40 cycles each, where the retained H2O uptake capacity of the framework was determined with relatively slower kinetics [191]. As displayed in Figure 40, H2N-UiO-66-Zr was observed to show a substantial and continuous loss of water uptake capacity, while the sorption behavior of H2N-MIL-125-Ti remained virtually unaltered during the short non-equilibrium cycles [191]. The water sorption kinetics for H2NMIL-125-Ti were observed to be faster than those for the H2N-UiO-66-Zr sample, while the weight loss of the dry sample as a result of the cycling experiments, was measured to be 2.7% for H2N-UiO-66-Zr, and 1.8% for H2N-MIL-125-Ti. Moreover, the loading lift and the sorption kinetics during the short, non-equilibrium cycles were observed to remain unchanged for H2NMIL-125-Ti throughout the hydrothermal cycle stress test, as can be evidenced from Figure 40. The PXRD patterns obtained before and after the test, shown in Figure 41, confirmed the retention of crystallinity and structural identity following the hydrothermal cycling treatment for both types of samples [191]. Conclusively, it was deduced from the study that aminefunctionalization is an effective strategy for improving the hydrothermal cyclic stability of the MIL-125-Ti framework, while a different PSM approach is required to improve the hydrothermal cyclic stability of UiO-66-Zr [191]. MIL-101(Cr) has been evaluated to adsorb the highest proportion of water molecules in the relative pressure range of 0.4< p / p 0 <0.54 [189], while a porous material is expected to perform in the 0.05< p / p 0 <0.4 range for satisfactory performance in a thermally-controlled 48

adsorption heat pump application. In order to modify the adsorption characteristics of MIL101(Cr), so as to achieve maximum adsorption capacity in the recommended relative pressure range, Anupam et al. [192] introduced hydrophilic nitro- or amino- functionalities into the framework using time-controlled PSM of the parent MOF, for intended use in thermally driven commercial adsorption chillers or heat pumps. Hydrothermal cycle stabilities over 40 ad/desorption cycles were investigated for four types of samples – MIL-101(Cr)-NH2, MIL101(Cr)-pNH2, MIL-101(Cr)-NO2, and MIL-101(Cr)-pNO2, where p stands for partial ligand amination/nitration. PXRD patterns, shown in Figure 42, confirmed that the intrinsic MIL101(Cr) framework structure and crystallinity were retained in all the tested samples, after the completion of 40 ad-/desorption cycles [192]. In order to evaluate the hydrothermal cyclic stability, the activated MOF samples were subjected to a continuous water adsorption and desorption cycle experiment between 40oC and 140oC, under a water vapor pressure of 5.2 kPa. The exchanged mass of water was followed over 40 repeated cycles for each sample, as shown in Figure 43 [192]. MIL-101(Cr)-NH2, MIL-101(Cr)-NO2, and MIL-101(Cr)-pNO2 were observed to exhibit an initial drop in the exchanged amount of water; however, from the fourth and the 15th cycle onwards, the desorbed and adsorbed mass of water were observed to remain constant for MIL-101(Cr)-NH2 and MIL-101(Cr)-NO2 respectively. For MIL-101(Cr)-pNO2, the mass of water was observed to decrease continuously over the designated period of 40 cycles. The partially aminated MIL-101(Cr)-pNH2 was observed to show the best water loading, according to mass and enthalpy measurements, for 40 subsequent water sorption cycles [192]. From the N2 adsorption isotherms obtained at 77 K, the amino-functionalized samples (MIL-101(Cr)-NH2 and MIL-101(Cr)-pNH2) were observed to retain their respective BET surface areas, and pore volumes at the end of the cycling experiment [192]. However, significant decreases in both the

49

surface areas and pore volumes were detected for samples MIL-101(Cr)-NO2 and MIL-101(Cr)pNO2 after cycle measurements, indicative of the fact that nitro-functionalized MIL-101(Cr) is

not recommended for use as adsorbent for commercial water-sorption-based heat transformation applications, over an extended period of time. 5. Summary and Conclusions

A comprehensive overview of the various physical and chemical factors which control the structural stability of MOFs in aqueous media, as well as ways to improve their hydrostability and hydrothermal cyclic stability have been presented. Table 6 summarizes the various MOFs that have been reported in this article to be structurally stable in one or more of the highlighted aqueous media, which include ambient moisture, room temperature water, boiling water, steam, aqueous acidic solutions, and aqueous basic solutions. From the published research literature summarized in the review, the following conclusions can be drawn concerning the structural stability of MOFs in aqueous media: 1. Among the MOFs reported to date, Zn-based MOFs, particularly IRMOF-1, have been observed to be most moisture-sensitive owing to their relatively weak metal-oxygen coordination bonds, which are vulnerable to hydrolysis and lead to the disruption of the open framework structure. 2. Chemical stability of carboxylate-based MOFs in aqueous media increases with increasing inertness of the central metal ion included in the framework. 3. Basicity of Ligand is the most important parameter which affects the structural stability of MOFs in aqueous media, since higher ligand basicity results in greater metal-ligand bond strength.

50

4. MOFs based on the highly basic pyrazole (pKa ≈ 19.8 ) and imidazole (pKa ≈ 18.6 ) ligands, in general, exhibit higher chemical resistivity to water than carboxylate-based MOFs. 5. Metal centers with lower coordination numbers are expected to result in shorter bond lengths, and have been suggested to result in relatively greater metal-ligand bond strengths. 6. MOFs containing 6-coordinate (usually octahedral) metal centers tend to be more stable than those containing 4-coordinate (usually tetrahedral) metal centers. 7. Metal centers or clusters with higher oxidation states result in relative higher stability towards reaction with water molecules. 8. Metal-ligand bond strength is a more contributing factor than coordination geometry, or valence, for M3+-containing MOFs than for M2+-containing ones (M = metal center). 9. The Zr6 cluster has been observed to be one of the most stable SBUs for MOF construction, and is largely responsible for exceptional hydrothermal stability of MOFs like PCN-222. 10. In-situ functionalization of the organic linker is the most effective strategy for improving the hydrostability and hydrothermal cyclic stability of MOFs for use in commercial applications since it is most time-efficient and cost-effective amongst the various methods discussed in the review, and also successfully acquires the required basicity of the ligand to strengthen the metal-ligand bond for achieving the desired level of structural stability. 11. Using catenation in combination with a pillaring strategy, it is possible to obtain waterstable MOFs even when the pillar ligand is characterized by relatively lower basicity.

51

Acknowledgements

The authors are highly indebted to the kind and continuous support extended by the Center of Research Excellence in Renewable Energy, King Fahd University of Petroleum and Minerals, in all stages of preparation of this review. References

[1]

A. U. Czaja, N. Trukhan, U. Müller, Chem. Soc. Rev. 38 (2009) 1284–1293.

[2]

U. Mueller, M. Schubert, F. Teich, H. Puetter, K. Schierle-Arndt, J. Pastré, J. Mater. Chem. 16 (2006) 626–636.

[3]

Z. Chen, S. Xiang, H. D. Arman, P. Li, S. Tidrow, D. Zhao, B. Chen, Eur. J. Inorg. Chem. 2010 (2010) 3745–3749.

[4]

F. Ma, S. Liu, D. Liang, G. Ren, C. Zhang, F. Wei, Z. Su, Eur. J. Inorg. Chem. 2010 (2010) 3756–3761.

[5]

J.-R. Li, Y. Ma, M. C. McCarthy, J. Sculley, J. Yu, H.-K. Jeong, P. B. Balbuena, H.-C. Zhou, Coord. Chem. Rev. 255 (2011) 1791–1823.

[6]

D. Liu, C. Zhong, J. Mater. Chem. 20 (2010) 10308–10318.

[7]

L. J. Murray, M. Dincă, J. R. Long, Chem. Soc. Rev. 38 (2009) 1294–1314.

[8]

J.-R. Li, R. J. Kuppler, H.-C. Zhou, Chem. Soc. Rev. 38 (2009) 1477–1504.

[9]

G. Férey, C. Serre, T. Devic, G. Maurin, H. Jobic, P. L. Llewellyn, G. De Weireld, A. Vimont, M. Daturi, J.-S. Chang, Chem. Soc. Rev. 40 (2011) 550–562.

[10]

S. K. Nune, P. K. Thallapally, B. P. McGrail, J. Mater. Chem. 20 (2010) 7623–7625.

[11]

D. Farrusseng, S. Aguado, C. Pinel, Angew. Chem. Int. Ed. 48 (2009) 7502–7513.

52

[12]

J. Lee, O. K. Farha, J. Roberts, K. A. Scheidt, S. T. Nguyen, J. T. Hupp, Chem. Soc. Rev. 38 (2009) 1450–1459.

[13]

L. Ma, C. Abney, W. Lin, Chem. Soc. Rev. 38 (2009) 1248–1256.

[14]

T. Ladrak, S. Smulders, O. Roubeau, S. J. Teat, P. Gamez, J. Reedijk, Eur. J. Inorg. Chem. 2010 (2010) 3804–3812.

[15]

W. Kleist, F. Jutz, M. Maciejewski, A. Baiker, Eur. J. Inorg. Chem. 2009 (2009) 3552– 3561.

[16]

M. D. Allendorf, C. A. Bauer, R. K. Bhakta, R. J. T. Houk, Chem. Soc. Rev. 38 (2009) 1330–1352.

[17]

C. Janiak, T. G. Scharmann, P. Albrecht, F. Marlow, R. Macdonald, J. Am. Chem. Soc. 118 (1996) 6307–6308.

[18]

H. A. Habib, A. Hoffmann, H. A. Hoeppe, G. Steinfeld, C. Janiak, Inorg. Chem. 48 (2009) 2166–2180.

[19]

M. Kurmoo, Chem. Soc. Rev. 38 (2009) 1353–1379.

[20]

B. Gil-Hernandez, P. Gili, J. K. Vieth, C. Janiak, J. Sanchiz, Inorg. Chem. 49 (2010) 7478–7490.

[21]

H. A. Habib, J. Sanchiz, C. Janiak, Inorg. Chim. Acta 362 (2009) 2452–2460.

[22]

C. Janiak, Dalton Trans. (2003) 2781–2804.

[23]

M. Rosseinsky, Microporous Mesoporous Mater. 73 (2004) 15–30.

[24]

G. Férey, Chem. Soc. Rev. 37 (2008) 191–214.

[25]

S. K. Henninger, F. Jeremias, H. Kummer, C. Janiak, Eur. J. Inorg. Chem. 2012 (2012) 2625–2634.

53

[26]

A. Rezk, R. Al-Dadah, S. Mahmoud, A. Elsayed, Int. J. Heat Mass Transfer 55 (2012) 7366–7374.

[27]

F. Jeremias, A. Khutia, S. K. Henninger, C. Janiak, J. Mater. Chem. 22 (2012) 10148– 10151.

[28]

J-Y Lee, C. Y. Tang, F. Huo, Scientific Reports 4, Article number: 3740.

[29]

Z. Hu, Y. Chen, J. Jiang, J. Chem. Phys. 134 (2011) 134705-1–134705-6.

[30]

J. Jiang, Mol. Simulat. 40 (2014) 516–536.

[31]

S. S. Kaye, A. Dailly, O. M. Yaghi, J. R. Long, J. Am. Chem. Soc. 129 (2007) 14176– 14177.

[32]

L. Huang, H. Wang, J. Chen, Z. Wang, J. Sun, D. Zhao, Y. Yan, Microporous Mesoporous Mater. 58 (2003) 105–114.

[33]

H. Jasuja, Y.-g. Huang, K. S. Walton, Langmuir 28 (2012) 16874–16880.

[34]

K. Tan, N. Nijem, P. Canepa, Q. Gong, J. Li, T. Thonhauser, Y. J. Chabal, Chem. Mater. 24 (2012) 3153–3167.

[35]

H. Jasuja, N. C. Burtch, Y.-g. Huang, Y. Cai, K. S. Walton, Langmuir 29 (2013) 633– 642.

[36]

T. Loiseau, C. Serre, C. Huguenard, G. Fink, F. Taulelle, M. Henry, T. Bataille, G. Férey, Chem.–Eur. J. 10 (2004) 1373–1382.

[37]

S. Surble, C. Serre, C. Mellot-Draznieks, F. Millange, G. Férey, Chem. Commun. (2006) 284–286.

[38]

J. A. Greathouse, M. D. Allendorf, J. Am. Chem. Soc. 128 (2006) 10678–10679.

[39]

T. J. Wu, L. J. Shen, M. Luebbers, C. H. Hu, Q.M. Chen, Z. Ni, R. I. Masel, Chem. Commun. 46 (2010) 6120–6122.

54

[40]

S. Hausdorf, J. Wagler, R. Mossig, F. Mertens, J. Phys. Chem. A 112 (2008) 7567–7576.

[41]

A. O. Yazaydin, A. I. Benin, S. A. Faheem, P. Jakubczak, J. J. Low, R. R. Willis, R. Q. Snurr, Chem. Mater. 21 (2009) 1425–1430.

[42]

J. Yang, A. Grzech, F. M. Mulderb, T. J. Dingemans, Chem. Commun. 47 (2011) 5244– 5246.

[43]

J. G. Nguyen, S. M. Cohen, J. Am. Chem. Soc. 132 (2010) 4560–4561.

[44]

P. Küsgens, M. Rose, S. Senkovska, H. Frçde, A. Henschel, S. Siegle, S. Kaskel, Microporous Mesoporous Mater. 120 (2009) 325–330.

[45]

K. Schröck, F. Schrçder, M. Heyden, R. A. Fischer, M. Havenith, Phys. Chem. Chem. Phys. 10 (2008) 4732–4739.

[46]

S. Paranthaman, F.-X. Coudert, A. H. Fuchs, Phys. Chem. Chem. Phys. 12 (2010) 8123– 8129.

[47]

S. Bourrelly, A. Vimont, G. Maurin, S. Devautour-Vinot, T. Devic, P. Horcajada, C. Serre, G. Clet, M. Daturi, Y. Filinchuk, G. Férey, P. L. Llewellyn, J. Am. Chem. Soc. 132 (2010) 9488–9498.

[48]

F. Gul-E-Noor, B. Jee, A. Pöppl, M. Hartmann, D. Himsl, M. Bertmer, Phys. Chem. Chem. Phys. 13 (2011) 7783–7788.

[49]

Y. Cheng, A. Kondo, H. Noguchi, H. Kajiro, K. Urita, T. Ohba, K. Kaneko, H. Kanoh, Langmuir 25 (2009) 4510 –4513.

[50]

M. Sabo, A. Henschel, H. Fröde, E. Klemm, S. Kaskel, J. Mater. Chem. 17 (2007) 3827– 3832.

[51]

Y. Li, R.T. Yang, Langmuir 23 (2007) 12937–12944.

55

[52]

J. J. Low, A. I. Benin, P. Jakubczak, J. F. Abrahamian, S. A. Faheem, R. R. Willis, J. Am. Chem. Soc. 131 (2009) 15834 –15842.

[53]

T.B. Čelič, M. Mazaj, G. Mali, V. Kaučič, N. Z. Logar, In: Proceedings of the 5th Serbian-Croatian-Slovenian Symposium on Zeolites, Zlatibor, Serbia, 30th May–2nd June, 2013.

[54]

M. Trzpit, M. Soulard, J. Patarin, N. Desbiens, F. Cailliez, A. Boutin, I. Demachy, A. H. Fuchs, Langmuir 23 (2007) 10131–10139.

[55]

F. Cailliez, G. Stirnemann, A. Boutin, I. Demachy, A. H. Fuchs, J. Phys. Chem. C 112 (2008) 10435–10445.

[56]

M. D. Toni, R. Jonchiere, P. Pullumbi, F.-X. Coudert, A. H. Fuchs, ChemPhysChem 13 (2012) 3497–3503.

[57]

I. J. Kang, N. A. Khan, E. Haque, S. H. Jhung, Chem. Eur. J. 17 (2011) 6437 – 6442.

[58]

J. H. Cavka, S. Jakobsen, U. Olsbye, N. Guillou, C. Lamberti, S. Bordiga, K. P. Lillerud, J. Am. Chem. Soc. 130 (2008) 13850−13851.

[59]

P. D. C. Dietzel, Y. Morita, R. Blom, H. Fjellvag, Angew. Chem., Int. Ed. 44 (2005) 6354−6358.

[60]

D. Ma, Y. Li, Z. Li, Chem. Commun., 47 (2011) 7377–7379.

[61]

K. A. Cychosz, A. J. Matzger, Langmuir 26 (2010) 17198−17202.

[62]

P. Küsgens, M. Rose, I. Senkovska, H. Fröde, A. Henschel, S. Siegle, S. Kaskel, Microporous Mesoporous Mater. 120 (2009) 325−330.

[63]

J.G. Nguyen, S.M. Cohen, J. Am. Chem. Soc. 132 (2010) 4560−4561.

[64]

C. Serre, Angew. Chem., Int. Ed. 51 (2012) 6048−6050.

56

[65]

J. Canivet, A. Fateeva, Y. Guo, B. Coasne, D. Farrusseng, Chem. Soc. Rev., (2014) DOI: 10.1039/C4CS00078A.

[66]

H. Taube, Chem. Rev. 50 (1952) 69–78.

[67]

A. Nimmermark, L. Öhrström, J. Reedijk, Z. Kristallogr. 228 (2013) 311–317.

[68]

J. Clayden, N. Greeves, S. Warren, P. Wothers, Organic Chemistry, second ed., Oxford University Press, Oxford, 2000.

[69]

G. Wulfsberg, Inorganic chemistry, University Science Books, Sausalito, Calif., 2000.

[70]

V. Colombo, S. Galli, H. J. Choi, G. D. Han, A. Maspero, G. Palmisano, N. Masciocchi, J. R. Long, Chem. Sci., 2 (2011) 1311–1319.

[71]

H. J. Choi, M. Dincă, A. Dailly, J. R. Long, Energy Environ. Sci., 3 (2010) 117–123.

[72]

C. R. Wade, T. Corrales-Sanchez, T. C. Narayan, M. Dincǎ, Energy Environ. Sci., 6 (2013) 2172–2177.

[73]

H. Jasuja, K. S. Walton, Dalton Trans., 42 (2013) 15421–15426.

[74]

Z. Liang, M. Marshall, A.L. Chaffee, Microporous Mesoporous Mater. 132 (2010) 305−310.

[75]

P.M. Schoenecker, C.G. Carson, H. Jasuja, C.J.J. Flemming, K.S. Walton, Ind. Eng. Chem. Res. 51 (2012) 6513−6519.

[76]

Q. Zheng, F. Yang, M. Deng, Y. Ling, X. Liu, Z. Chen, Y. Wang, L. Weng, Y. Zhou, Inorg. Chem. 52 (2013) 10368–10374.

[77]

C. S. Hawes, P. E. Kruger, Cryst. 4 (2014) 32–41.

[78]

C. Serre, F. Millange, C. Thouvenot, M. Noguès, G. Marsolier, D. Louër, G. Férey, J. Am. Chem. Soc. 124 (2002) 13519−13526.

57

[79]

G. Férey, C. Serre, C. Mellot-Draznieks, F. Millange, S. Surblé, J. Dutour, I. Margiolaki, Angew. Chem., Int. Ed. 43 (2004) 6296−6301.

[80]

M. Dan-Hardi, C. Serre, T. Frot, L. Rozes, G. Maurin, C. Sanchez, G. Férey, J. Am. Chem. Soc. 131 (2009) 10857−10859.

[81]

N.L. Rosi, J. Kim, M. Eddaoudi, B. Chen, M. O'Keeffe, O.M. Yaghi, J. Am. Chem. Soc. 127 (2005) 1504–1518.

[82]

S.S.Y. Chui, S.M.F. Lo, J.P.H. Charmant, A.G. Orpen, I.D. Williams, Science 283 (1999) 1148–1150.

[83]

J.B. DeCoste, G.W. Peterson, B.J. Schindler, K.L. Killops, M.A. Browe, J.J. Mahl, J. Mater. Chem. A 1 (2013) 11922–11932.

[84]

J.B. Decoste, G.W. Peterson, M.W. Smith, C.A. Stone, C. R. Willis, J. Am. Chem. Soc. 134 (2012) 1486–1489.

[85]

T. G. Glover, G. W. Peterson, B. J. Schindler, D. Britt, O.M. Yaghi, Chem. Eng. Sci. 66 (2011) 163–170.

[86]

D. Britt, D. Tranchemontagne, O.M. Yaghi, Proc. Natl. Acad. Sci. U.S.A. 105 (2008) 11623–11627.

[87]

S.R. Caskey, A.G. Wong-Foy, A.J. Matzger, J. Am. Chem. Soc. 130 (2008) 10870– 10871.

[88]

D. Britt, H. Furukawa, B. Wang, T.G. Glover, O.M. Yaghi, Proc. Natl. Acad. Sci. U. S. A. 106 (2009) 20637–20640.

[89]

A.C. Kizzie, A.G. Wong-Foy, A. J. Matzger, Langmuir 27 (2011) 6368–6373.

[90]

J. Liu, A.I. Benin, A.M.B. Furtado, P. Jakubczak, R.R. Willis, M.D. LeVan, Langmuir 27 (2011) 11451–11456.

58

[91]

M. Kandiah, M.H. Nilsen, S. Usseglio, S. Jakobsen, U. Olsbye, M. Tilset, C. Larabi, E. A. Quadrelli, F. Bonino, K.P. Lillerud, Chem. Mater. 22 (2010) 6632–6640.

[92]

L. Valenzano, B. Civalleri, S. Chavan, S. Bordiga, M.H. Nilsen, S. Jakobsen, K.P. Lillerud, C. Lamberti, Chem. Mater. 23 (2011) 1700–1718.

[93]

J.B. DeCoste, G.W. Peterson, H. Jasuja, T.G. Glover, Y.-g. Huang, K.S. Walton, J. Mater. Chem. A, 1 (2013) 5642–5650.

[94]

A. Schaate, P. Roy, A. Godt, J. Lippke, F. Waltz, M. Wiebcke, P. Behrens, Chem.–Eur. J. 17 (2011) 6643–6651.

[95]

M. Kim, S. M. Cohen, CrystEngComm 14 (2012) 4096–4104.

[96]

M. Kim, J.F. Cahill, H. Fei, K.A. Prather, S.M. Cohen, J. Am. Chem. Soc. 134 (2012) 18082–18088.

[97]

M. Kim, J.F. Cahill, Y. Su, K.A. Prather, S.M. Cohen, Chem. Sci. 3 (2012) 126–130.

[98]

W. Mori, F. Inoue, K. Yoshida, H. Nakayama, Takamiza-wa, M. Kishita, Chem. Lett. 26 (1997) 1219–1220.

[99]

L. Deakin, A.M. Arif, J.S. Miller, Inorg. Chem. 38 (1999) 5072–5077.

[100] K. Seki, Chem. Commun. 30 (2001) 1496–1497. [101] K. Seki, W. Mori, J. Phys. Chem. B. 106 (2002) 1380–1385. [102] D.N. Dybtsev, H. Chun, K. Kim, Angew. Chem. Int. Ed. 43 (2004) 5033–5036. [103] J.Y. Lee, D.H. Olson, L. Pan, T.J. Emge, J. Li, Adv. Funct. Mater. 17 (2007) 1255–1262. [104] J.Y. Lee, L. Pan, X.Y. Huang, T.J. Emge, J. Li, Adv. Func. Mater. 21 (2011) 993–998. [105] Y.G. Zhao, H.H. Wu, T.J. Emge, Q.H. Gong, N. Nijem, Y.J. Chabal, L.Z. Kong, D.C. Langreth, H. Liu, H.P. Zeng, J. Li, Chem. Euro. J. 17 (2011) 5101–5109.

59

[106] M. Haouas, C. Volkringer, T. Loiseau, G. Férey, F. Taulelle, Chem. Euro. J. 15 (2009) 3139–3146. [107] W. Clegg, D.R. Harbron, C.D. Homan, P.A. Hunt, I.R. Little, B.P. Straughan, Inorg. Chim. Acta 186 (1991) 51---60. [108] B. Chen, C. Liang, J. Yang, D.S. Contreras, Y.L. Clancy, E.B. Lobkovsky, O.M. Yaghi, S. Dai, Angew. Chem., Int. Ed. 45 (2006) 1390–1393. [109] I.V. Pletnev, V.V. Zernov, Anal. Chim. Acta 455 (2002) 131–142. [110] K.S. Park, Z. Ni, A.P. Cote, J.Y. Choi, R. Huang, F.J. Uribe-Romo, H.K. Chae, M. O'Keeffe, O.M. Yaghi, Proc. Natl. Acad. Sci. U.S.A. 103 (2006) 10186–10191. [111] L. Bellarosa, J.M. Castillo, T. Vlugt, S. Calero, N. López, Chem.–Eur. J. 18 (2012) 12260–12266. [112] M. Dincǎ, A.F. Yu, J.R. Long, J. Am. Chem. Soc. 128 (2006) 8904–8913. [113] N. Masciocchi, S. Galli, V. Colombo, A. Maspero, G. Palmisano, B. Seyyedi, C. Lamberti, S. Bordiga, J. Am. Chem. Soc. 132 (2010) 7902–7904. [114] A. Demessence, J. R. Long, Chem.–Eur. J. 16 (2010) 5902–5908. [115] J.-P. Zhang, Y.-B. Zhang, J.-B. Lin, X.-M. Chen, Chem. Rev. 112 (2012) 1001–1033. [116] F.G. Bordwell, Acc. Chem. Res. 21 (1988) 456–463. [117] G. Nickerl, M. Leistner, S. Helten, V. Bon, I. Senkovska, S. Kaskel, Inorg. Chem. Front. 1 (2014) 325–330. [118] V. Bon, V. Senkovskyy, I. Senkovska, S. Kaskel, Chem. Commun. 48 (2012) 8407– 8409. [119] P. Horcajada, S. Surble, C. Serre, D.-Y. Hong, Y.-K. Seo, J.-S. Chang, J.-M. Greneche, I. Margiolaki, G. Férey, Chem. Commun. (2007) 2820–2822.

60

[120] C. Volkringer, D. Popov, T. Loiseau, G.R. Férey, M. Burghammer, C. Riekel, M. Haouas, F. Taulelle, Chem. Mater. 21 (2009) 5695–5697. [121] A. Lieb, H. Leclerc, T. Devic, C. Serre, I. Margiolaki, F. Mahjoubi, J.S. Lee, A. Vimont, M. Daturi, J.-S. Chang, Microporous Mesoporous Mater. 157 (2012) 18–23. [122] J. Seo, R. Matsuda, H. Sakamoto, C. Bonneau, S. Kitagawa, J. Am. Chem. Soc. 131 (2009) 12792–12800. [123] P. Nugent, Y. Belmabkhout, S.D. Burd, A.J. Cairns, R. Luebke, K. Forrest, T. Pham, S. Ma, B. Space, L. Wojtas, M. Eddaoudi, M.J. Zaworotko, Nature 495 (2013) 80–84. [124] C.A. Kent, D. Liu, T.J. Meyer, W. Lin, J. Am. Chem. Soc. 134 (2012) 3991–3994. [125] E.-Y. Choi, P.M. Barron, R.W. Novotny, H.-T. Son, C. Hu, W. Choe, Inorg. Chem. 48 (2008) 426–428. [126] C.Y. Lee, O.K. Farha, B.J. Hong, A.A. Sarjeant, S.T. Nguyen, J.T. Hupp, J. Am. Chem. Soc. 133 (2011) 15858–15861. [127] D. Feng, Z.-Y. Gu, J.-R. Li, H.-L. Jiang, Z. Wei, H.-C. Zhou, Angew. Chem., Int. Ed. 51 (2012) 10307–10310. [128] H.-L. Jiang, D. Feng, K. Wang, Z.-Y. Gu, Z. Wei, Y.-P. Chen, H.-C. Zhou, J. Am. Chem. Soc. 135 (2013) 13934–13938. [129] D. Feng, H.-L. Jiang, Y.-P. Chen, Z.-Y. Gu, Z. Wei, H.-C. Zhou, Inorg. Chem. 52 (2013) 12661–12667. [130] D. Feng, W.-C. Chung, Z. Wei, Z.-Y. Gu, H.-L. Jiang, Y.-P. Chen, D.J. Darensbourg, H.C. Zhou, J. Am. Chem. Soc. 135 (2013) 17105–17110. [131] A. Schaate, P. Roy, T. Preuße, S.J. Lohmeier, A. Godt, P. Behrens, Chem. Eur. J. 17 (2011) 9320–9325.

61

[132] Y.-L. Zhao, L. Liu, W. Zhang, C.-H. Sue, Q. Li, O.S. Miljanić, O.M. Yaghi, J.F. Stoddart, Chem. Eur. J. 15 (2009) 13356–13380. [133] T.A. Makal, X. Wang, H.-C. Zhou, Cryst. Growth Des. 13 (2013) 4760−4768. [134] D.-S. Zhang, Z. Chang, Y.-F. Li, Z.-Y. Jiang, Z.-H. Xuan, Y.-H. Zhang, J.-R. Li, Q. Chen, T.-L. Hu, X.-H. Bu, Scientific Reports; 11/22/2013, p1. [135] J.M. Taylor, R. Vaidhyanathan, S.S. Iremonger, G.K.H. Shimizu, J. Am. Chem. Soc. 134 (2012) 14338−14340. [136] M. Dincǎ, J.R. Long, Angew. Chem., Int. Ed. 47 (2008) 6766–6779. [137] A.R. Millward, O.M. Yaghi, J. Am. Chem. Soc. 127 (2005) 17998–17999. [138] H. Furukawa, M.A. Miller, O.M. Yaghi, J. Mater. Chem. 17 (2007) 3197–3204. [139]

S. Ma, D. Sun, J.M. Simmons, C.D. Collier, D. Yuan, H.-C. Zhou, J. Am. Chem. Soc. 130 (2008) 1012–1016.

[140] R.E. Morris, P.S. Wheatley, Angew. Chem., Int. Ed. 47 (2008) 4966–4981. [141] A. Demessence, D.M. D’Alessandro, M.L. Foo, J.R. Long, J. Am. Chem. Soc. 131 (2009) 8784–8786. [142] E. Quartapelle-Procopio, F. Linares, C. Montoro, V. Colombo, A. Maspero, E. Barea, J.A.R. Navarro, Angew. Chem., Int. Ed. 49 (2010) 7308–7311. [143] S. Horike, M. Dincǎ, K. Tamaki, J.R. Long, J. Am. Chem. Soc. 130 (2008) 5854–5855. [144] R. Wang, X.-Y. Dong, H. Xu, R.-B. Pei, M.-L. Ma, S.-Q. Zang, H.-W. Hou, T. C. W. Mak, Chem. Commun. (2014) DOI: 10.1039/C4CC03268C. [145] J.-M. Zhou, W. Shi, N. Xu, P. Cheng, Inorg. Chem. 52 (2013) 8082–8090. [146] J.-M. Zhou, W. Shi, N. Xu, P. Cheng, Inorg. Chem. 52 (2013) 8082–8090 (Supporting information).

62

[147] Y. Liu, K. Mo, Y. Cui, Inorg. Chem. 52 (2013) 10286–10291. [148] M. Gustafsson, A. Bartoszewicz, B. Martín-Matute, J. Sun, J. Grins, T. Zhao, Z. Li, G. Zhu, X. Zou, Chem. Mater. 22 (2010) 3316–3322. [149] Y. Liu, K. Mo, Y. Cui, Inorg. Chem. 52 (2013) 10286–10291 (Supporting information). [150] S.M. Cohen, Chem. Rev. 112 (2012) 970–1000. [151] A.D. Burrows, CrystEngComm 13 (2011) 3623–3642. [152] Z. Wang, S.M. Cohen, Chem. Soc. Rev. 38 (2009) 1315–1329. [153] K.K. Tanabe, Z. Wang, S.M. Cohen, J. Am. Chem. Soc. 130 (2008) 8508–8517. [154] Z. Wang, K.K. Tanabe, S.M. Cohen, Inorg. Chem. 48 (2009) 296–306. [155] Y. Huang, W. Qin, Z. Li, Y. Li, Dalton Trans. 41 (2012) 9283–9285. [156] J. Aguilera-Sigalat, D. Bradshaw, Chem. Commun. 50 (2014) 4711—4713 (Supporting information). [157] J. H. Im, N. Ko, S. J. Yang, H. J. Park, J. Kim, C. R. Park, New J.Chem. 38 (2014) 2752– 2755. [158] M. Kondo, T. Okubo, A. Asami, S. Noro, T. Yoshitomi, S. Kitagawa, T. Ishii, H. Matsuzaka, K. Seki, Angew. Chem., Int. Ed. 38 (1999) 140–143. [159] S.-H. Cho, B. Ma, S.T. Nguyen, J.T. Hupp, T.E. Albrecht-Schmitt, Chem. Commun. (2006) 2563–2565. [160] B. Chen, S. Ma, F. Zapata, F.R. Fronczek, E.B. Lobkovsky, H.-C. Zhou, Inorg. Chem. 46 (2007) 1233–1236. [161] Z. Lu, H. Xing, R. Sun, J. Bai, B. Zheng, Y. Li, Cryst. Growth Des. 12 (2012) 1081−1084.

63

[162] H. Liu, Y. Zhao, Z. Zhang, N. Nijem, Y.J. Chabal, X. Peng, H. Zeng, J. Li, Chem. Asian J. 8 (2013) 778–785 (Supporting information). [163] H. Liu, Y.G. Zhao, Z.J. Zhang, N. Nijem, Y.J. Chabal, H.P. Zeng, J. Li, Adv. Funct. Mater. 21 (2011) 4754–4762. [164] L.-H. Xie, M.P. Suh, Chem.–Eur. J. 17 (2011) 13653–13656. [165] H. Chun, D.N. Dybtsev, H. Kim, K. Kim, Chem.–Eur. J. 11 (2005) 3521–3529. [166] B. Chen, M. Eddaoudi, S.T. Hyde, M. O'Keeffe, O.M. Yaghi, Science 291 (2001) 1021– 1023. [167] B. Chen, M. Eddaoudi, T.M. Reineke, J.W. Kampf, M. O'Keeffe, O.M. Yaghi, J. Am. Chem. Soc. 122 (2000) 11559–11560. [168] H. Li, M. Eddaoudi, M. O'Keeffe, O.M. Yaghi, Nature 402 (1999) 276–279. [169] S. Hilal, S.W. Karickhoff, L.A. Carreira, Quant. Struc. Act. Rel. 14 (1995) 348–355. [170] J. Zhang, L. Wojtas, R.W. Larsen, M. Eddaoudi, M.J. Zaworotko, J. Am. Chem. Soc. 131 (2009) 17040–17041. [171] J.B. Decoste, G.W. Peterson, J. Vis. Exp. 80 (2013) DOI: 10.3791/51175. [172] S.J. Yang, C.R. Park, Adv. Mater. 24 (2012) 4010–4013. [173] L.M. Lanni, R.W. Tilford, M. Bharathy, J.J. Lavigne, J. Am. Chem. Soc. 133 (2011) 13975–13983. [174] S.J. Yang, J.Y. Choi, H.K. Chae, J.H. Cho, K.S. Nahm, C.R. Park, Chem. Mater. 21 (2009) 1893–1897. [175] B. Liu, H. Shioyama, T. Akita, Q. Xu, J. Am. Chem. Soc. 130 (2008) 5390–5391. [176] H.L. Jiang, B. Liu, Y.Q. Lan, K. Kuratani, T. Akita, H. Shioyama, F. Zong, Q. Xu, J. Am. Chem. Soc. 133 (2011) 11854–11857.

64

[177] S.J. Yang, T. Kim, J.H. Im, Y.S. Kim, K. Lee, H. Jung, C.R. Park, Chem. Mater. 24 (2012) 464–470. [178] H. Li, W. Shi, K. Zhao, H. Li, Y. Bing, P. Cheng, Inorg. Chem. 51 (2012) 9200−9207. [179] J.T. Han, S.Y. Kim, J.S. Woo, G.-W. Lee, Adv. Mater. 20 (2008) 3724–3727. [180] B. Panella, M. Hirscher, H. Pütter, U. Müller, Adv. Funct. Mater. 16 (2006) 520–524. [181] C.-M. Wu, M. Rathi, S.P. Ahrenkiel, R.T. Koodali, Z. Wang, Chem. Commun. 49 (2013) 1223–1225. [182] A. Fillinger, B.A. Parkinson, J. Electrochem. Soc. 146 (1999) 4559–4564. [183] L.-B. Sun, J.-R. Li, W. Lu, Z.-Y. Gu, Z. Luo, H.-C. Zhou, J. Am. Chem. Soc. 134 (2012) 15923–15928. [184] D. Buso, K.M. Nairn, M. Gimona, A.J. Hill, P. Falcaro, Chem. Mater. 23 (2011) 929– 934. [185] P. Pachfule, B.K. Balan, S. Kurungot, R. Banerjee, Chem. Commun. 48 (2012) 2009– 2011. [186] X. Chen, P. Lukaszczuk, C. Tripisciano, M.H. Rümmeli, J. Srenscek-Nazzal, I. Pelech, R.J. Kalenczuk, E. Borowiak-Palen, Phys. Status Solidi B 247 (2010) 2664–2668. [187] S.K. Henninger, H.A. Habib, C. Janiak, J. Am. Chem. Soc. 131 (2009) 2776–2777. [188] A. Rezk, R. Al-Dadah, S. Mahmoud, A. Elsayed, P. I. Mech. Eng. C-J. Mec. 227 (2013) 992–1005. [189] J. Ehrenmann, S.K. Henninger, C. Janiak, Eur. J. Inorg. Chem. (2011) 471–474. [190] C. Janiak, S.K. Henninger, Chimia 67 (2013) 419–424. [191] F. Jeremias, V. Lozan, S.K. Henninger, C. Janiak, Dalton Trans. 42 (2013) 15967–15973.

65

[192] A. Khutia, H.U. Rammelberg, T. Schmidt, S.K. Henninger, C. Janiak, Chem. Mater. 25 (2013) 790−798. [193] G. Akiyama, R. Matsuda, H. Sato, A. Hori, M. Takata, S. Kitagawa, Microporous Mesoporous Mater. 157 (2012) 89–93. [194] S.J. Yang, C.R. Park, Adv. Mater. 24 (2012) 4010–4013 (Supporting information). [195] S.J. Yang, J.Y. Choi, H.K. Chae, J.H. Cho, K.S. Nahm, C.R. Park, Chem. Mater. 21 (2009) 1893–1897 (Supporting information).

66

Figure 1 – Intermediate hydration states of a Zn2+ coordination sphere in IRMOF-0h with zero or one bound water molecule [Reprinted with permission from ref. 56 (Copyright 2012, WileyVCH Verlag)].

(a)

(b)

Figure 2 – (a) Factors controlling the structural stability of MOFs in aqueous media, and (b) Methods used to improve the hydrostability and hydrothermal cyclic stability of MOFs (PECVD = Plasma Enhanced Chemical Vapor Deposition).

67

Figure 3 – Organic ligands employed in the synthesis of pillared MOFs [Reprinted with

permission from ref. 73 (Copyright 2013, Royal Society of Chemistry)].

Figure 4 – Cu-BTC (top), Mg-MOF-74 (middle), and UiO-66 (bottom) viewed along two

different axes to show the various pore structures of Cu-BTC and UiO-66, and the 1-D pores of Mg-MOF-74. Color scheme: C (gray), H (white), O (red), Cu (orange), Mg (purple), and Zr (light blue) [Reprinted with permission from ref. 83 (Copyright 2013, Royal Society of Chemistry)]. 68

(a)

(b)

Figure 5 – SBUs for M3+ (M = Al or Cr) containing MOFs: (a) Chains of corner-sharing

octahedra in Cr- or Al-MIL-53, and (b) Al8(OH) 159+ octamers in MIL-110 [Reprinted with permission from ref. 52 (Copyright 2009, American Chemical Society)].

Figure 6 – PXRD patterns of DUT-51(Zr): (a) calculated from single-crystal X-ray

structure, (b) as-synthesized, (c) activated at 120oC, and (d) after soaking in water for 12 hours at room temperature [Reprinted with permission from ref. 118 (Copyright 2012, Royal Society of Chemistry)].

69

Figure 7 – Crystal structure and underlying network topology of PCN-222(Fe). The Fe-TCPP

(blue square in a)) is connected to four 8-connected Zr6 clusters (light orange cuboid in b)), with a twisted angle, to generate a 3-D network in Kagome-like topology (d) and e)) with 1-D large channels (green pillar in c)). Color scheme: Zr (black spheres), C (gray), O (red), N (blue), Fe (orange). H atoms are omitted for clarity [Reprinted with permission from ref. 127 (Copyright 2013, Wiley-VCH Verlag)].

70

Figure 8 – PXRD patterns for PCN-222(Fe) at 77 K, showing framework stability upon

treatment with water, boiling water, 2M, 4M, 8M, and concentrated HCl [Reprinted with permission from ref. 127 (Copyright 2013, Wiley-VCH Verlag)].

71

Scheme 1 – Syntheses of rodlike dicarboxylic acids HO2C[PE-P(R1,R2)-EP]CO2H (1b–8b) from

the diesters AlkO2C[PE-P(R1,R2)-EP]CO2Alk (1a–8a) [Reprinted with permission from ref. 131 (Copyright 2011, Wiley-VCH Verlag)].

72

Figure 9 – Alternating pairs of edge-sharing tetrahedral, and edge-sharing octohedra in chain

formation of MOF-69C framework [Reprinted with permission from ref. 52 (Copyright 2009, American Chemical Society)].

Figure 10 – Chains of edge-sharing Zn octahedra in MOF-74 [Reprinted with permission

from ref. 52 (Copyright 2009, American Chemical Society)].

73

(a)

(b)

Figure 11 – PXRD patterns of (a) methyl-modified MOF-5, and (b) 2,5-dimethyl-modified

MOF-5 before (black) and after (red) 4-day exposure to ambient air [Reprinted with permission from ref. 42 (Copyright 2011, Royal Society of Chemistry)].

74

Scheme 2 – Synthesis and structures of ligand precursors H4TPTC-OR ((i) K2CO3, DMF, and

RBr; (ii) ICI and MeOH; (iii) diethyl 5-(4,4,5,5-tetramethyl-,3,2-dioxaborolan-2-yl)-1,3benzene-dicarboxylate, NaHCO3, CsF, Pd(PPH3)4, 1,2-dimethoxyethane, and H2O; (iv) 1. MeOH, THF, and KOH 2. H3O+) [Reprinted with permission from ref. 133 (Copyright 2013, American Chemical Society)].

75

Figure 12 – PXRD patterns for Cu2TPTC-OR MOFs (as: as-synthesized, ac: activated), after

being exposed to atmosphere after activation for 1 h, and after being exposed to atmosphere (50% RH) after activation for 16 days, compared to the PXRD pattern for NOTT-106 (simulated from single-crystal data) [Reprinted with permission from ref. 133 (Copyright 2013, American Chemical Society)].

76

Figure 13 – Dry (a), c), e)) and wet (b), d), f)) powder samples of Cu2TPTC-OEt, −OnPr, and

−OnHex, respectively [Reprinted with permission from ref. 133 (Copyright 2013, American Chemical Society)].

(a)

(b)

Figure 14 – PXRD comparison of (a) Banasorb-22 and (b) IRMOF-1, before (black curve) and

after (red curve) water vapor exposure [Reprinted with permission from ref. 39 (Copyright 2010, Royal Society of Chemistry)].

77

Figure 15 – General route for the synthesis of TKL MOFs [Reprinted with permission

from ref. 134 (Copyright 2013, Nature Publishing Group)].

Figure 16 – H2O adsorption isotherms (blue), and the isosteric heat of adsorption of H2O (red)

for CALF-25. Fits to virial models are shown by dashed lines [Reprinted with permission from ref. 135 (Copyright 2012, American Chemical Society)].

78

Figure 17 – PXRD patterns for the Ni3(BTP)2 framework after treatment in water, acids, and

base for two weeks at 100oC [Reprinted with permission from ref. 70 (Copyright 2011, Royal Society of Chemistry)].

Figure 18 – PXRD patterns for the Cu3(BTP)2 framework during treatment in water for 14 days

at 100oC (top), and transformation into Cu3(BTP)2.6H2O after treatment in an acidic or a basic solution (bottom) [Reprinted with permission from ref. 70 (Copyright 2011, Royal Society of Chemistry)].

79

Figure 19 – PXRD patterns of Zn3(BTP)2, after treatment in water, acid, or base for various durations at various temperatures [Reprinted with permission from ref. 70 (Copyright 2011, Royal Society of Chemistry)].

(a)

(b)

Figure 20 – PXRD patterns of (a) [Eu2(BPDC)(BDC)2(H2O)2]n, and (b) [Tb2(BPDC)(BDC)2(H2O)2]n in as-synthesized states and after immersion in 0.01M aqueous NaX solutions for 2 days (X = F-, Cl-, Br-, and I-) [Reprinted with permission from ref. 146 (Copyright 2013, American Chemical Society)].

80

Figure 21 – PXRD patterns of [Nd(trimesate)] MOF in as-synthesized and activated

states, and after immersion in boiling water, boiling methanol, and boiling benzene for 1 week [Reprinted with permission from ref. 149 (Copyright 2013, American Chemical Society)].

81

(a)

(b)

Figure 22 – (a) PXRD pattern of UiO-66-N=N-ind1, after soaking for 2 h at pH 11 (red), and the

empty sample holder (black), and (b) Nitrogen sorption isotherms (77 K), after soaking for 2 h of UiO-66-NH2 at pH 9 (black), UiO-66-N=N-ind1h at pH 11 (red), and UiO-66-N=N-ind3h at pH 12 (blue) [Reprinted with permission from ref. 156 (Copyright 2014, Royal Society of Chemistry)].

82

(a)

(b)

Figure 23 – PXRD scans of (a) unmodified MIL-125, and (b) methyl-modified MIL-125 (black: as-synthesized, red: water-immersed for 1 hour, blue: water-immersed for 2 hours, green: waterimmersed for 4 hours) [Reprinted with permission from ref. 157 (Copyright 2014, Royal Society of Chemistry)].

Figure 24 – Schematic illustration of the evolutionary approach used to synthesize four types of MOF complexes starting from simple-layered and ending at pillar-layered structure [Reprinted with permission from ref. 161 (Copyright 2012, American Chemical Society)]. 83

Figure 25 – PXRD patterns for MOF complex synthesized with simple layers following

treatment in room temperature water, and boiling water for various durations [Reprinted with permission from ref. 161 (Copyright 2012, American Chemical Society)].

84

(a)

(b)

Figure 26 – PXRD patterns for compounds (a) [Zn3(bpdc)3( 2,2′ -dmbpy)].DMFx-(H2O)y, and (b)

[Zn3(bpdc)3(bpy)].DMF4-(H2O), after exposure to water vapor (323 K) (bpdc = biphenyl- 4,4′ dicarboxylic acid; z, z ′ -dmbpy = z, z ′ -dimethyl-4, 4 ′ -bypyridine; bpy = 4,4′ -bypyridine). From bottom to top: as-synthesized sample; after exposure for 1 day; after exposure for 5 days; after exposure for 10 days [Reprinted with permission from ref. 162 (Copyright 2013, Wiley-VCH Verlag)].

85

Figure 27 – Illustration of a) non-catenated pillared layer frameworks (DMOF, DMOF-TM,

MOF-508-TM), synthesized from dicarboxylate (red) and pillar ligand (green) (blue corners are metal nodes), and b) two-fold catenation in pillared layer frameworks (MOF-508) (Black and white represents two different frameworks) [Reprinted with permission from ref. 73 (Copyright 2013, Royal Society of Chemistry)].

Figure 28 – Two MOF-14 frameworks (blue and red), interwoven about a P-minimal surface

without intersecting the surface [Reprinted with permission from ref. 166 (Copyright 2001, AAAS, http://www.sciencemag.org)].

86

Figure 29 – PXRD patterns for activated, water-exposed, and regenerated MOF-508 [Reprinted

with permission from ref. 73 (Copyright 2013, Royal Society of Chemistry)].

Figure 30 – Snapshots of Cu-BTC dispersed in water (top, left), and C2F6 plasma-treated Cu-

BTC repelling and floating on top of water (bottom, left). Contact angle images of Cu-BTC (top, right), and C2F6 plasma-treated Cu-BTC (bottom, right), with a 2 µl droplet of water [Reprinted with permission from ref. 171 (Copyright 2013, Journal of Visualized Experiments)].

87

Figure 31 – PXRD patterns of Cu-BTC (black, bottom), Cu-BTC aged at 45 °C and 100% RH

for 3 days (blue, middle), and C2F6 plasma-treated Cu-BTC aged at 45 °C and 100% RH for 3 days (red, top) [Reprinted with permission from ref. 171 (Copyright 2013, Journal of Visualized Experiments)].

(a)

(b)

Figure 32 – PXRD patterns for (a) Cu-BTC, and (b) perfluorohexane plasma-treated Cu-BTC,

after immersion in water at room temperature for 0, 4, and 24 h [Reprinted with permission from ref. 84 (Copyright 2012, American Chemical Society)].

88

(a)

(b)

(c)

(d)

Figure 33 – Schematic representations of (a) untreated IRMOF-1, (b) thermally treated IRMOF-

1 with amorphous carbon coating, (c) thicker coating after prolonged treatment, and (d) formation of ZnO nanoparticles@amorphous carbon at higher temperatures. The corresponding XRD patterns are shown on the right [Reprinted with permission from ref. 172 (Copyright 2012, Wiley-VCH Verlag)].

89

Figure 34 – Snapshots of (a) untreated IRMOF-1 immediately after immersion in water, and (b) thermally-modified IRMOF-1 after immersion in water for 2 hours [Reprinted with permission from ref. 172 (Copyright 2012, Wiley-VCH Verlag)].

(a)

(b)

Figure 35 – PXRD patterns of (a) untreated IRMOF-1 and (b) thermally-modified IRMOF-1 after exposure to ambient conditions (black: before exposure, red: after 5-day exposure, and blue: after 14-day exposure) [Reprinted with permission from ref. 172 (Copyright 2012, WileyVCH Verlag)].

90

Figure 36 – (a) PXRD patterns of as synthesized MOF-5, Ni13-MOF-5, Ni22-MOF-5, and the

simulated MOF-5, and PXRD patterns, after exposure to static air conditions (25 °C and 30−37% RH), for (b) MOF-5, (c) Ni13-MOF-5, and (d) Ni22-MOF-5 [Reprinted with permission from ref. 178 (Copyright 2012, American Chemical Society)].

91

Figure 37 – PXRD patterns for (a) MOF-5, and (b) MOFMC, exposed to static air conditions

(23 °C and 33% RH), for <5 min, 2 h, 12 h, 36 h, and 168 h [Reprinted with permission from ref. 174 (Copyright 2009, American Chemical Society)].

Figure 38 – Proposed synthetic scheme for MOF-5@SBA-15 nanocomposite [Reprinted with

permission from ref. 181 (Copyright 2013, Royal Society of Chemistry)].

92

(a)

(b)

Figure 39 – Comparison of PXRD patterns of (a) MOF-5, and (b) MOF-5@SBA-15 nanocomposite, as a function of air exposure time from 0 hours to 7 days [Reprinted with permission from ref. 181 (Copyright 2013, Royal Society of Chemistry)].

(a)

(b)

Figure 40 – Hydrothermal cycle stress test plots, with temperature profile and load signal, for (a) H2N-UiO-66, and (b) H2N-MIL-125 frameworks [Reprinted with permission from ref. 191 (Copyright 2013, Royal Society of Chemistry)].

93

Figure 41 – PXRD patterns (Cu-Kα radiation) of H2N-UiO-66 (top, red), and H2N-MIL-125

(bottom, blue). (a, d): as simulated, (b, e): before hydrothermal cycling treatment, and (c, f ): after 40 water ad-/desorption cycles (cf. Figure 40) [Reprinted with permission from ref. 191 (Copyright 2013, Royal Society of Chemistry)].

Figure 42 – PXRD patterns of functionalized MIL-101(Cr) samples [Reprinted with permission

from ref. 192 (Copyright 2013, American Chemical Society)].

94

Figure 43 – Change of total mass variation in functionalized MIL-101(Cr) samples, during water

desorption and adsorption over 40 cycles ((1) = MIL-101(Cr)-NH2, (2) = MIL-101(Cr)-pNH2, (3) = MIL-101(Cr)-NO2, (4) = MIL-101(Cr)-pNO2) [Reprinted with permission from ref. 192 (Copyright 2013, American Chemical Society)].

95

Table 1 – Approaches adopted to improve the hydrostability of MOF-5 in various

aqueous media (RTW = room temperature water, BW = boiling water, AM = ambient moisture, S = steam, AAS = aqueous acidic solutions, ABS = aqueous basic solutions). Media in which modified MOF-5 Approach to improve hydrostability of

is hydrostable

MOF-5

(AM, RTW, BW, S, AAS, ABS)

Reference

Functionalization of the BDC linker with one or more hydrophobic functionalities

AM

[42]

RTW and AM

[43]

RTW and AM

[172]

AM

[174]

AM

[178]

AM

[181]

Post-synthetic modification of the BDC linker with hydrophobic moieties like alkyl chains Coating the MOF-5 structure with a more hydrophobic material Incorporation of second-phase hydrophobic nanoparticles Partial exchange of Zn(II) ions in the MOF-5 framework with other transition metal ions Confinement of MOF-5 particles in a waterresistant matrix

96

Table 2 – Comparison of basicity (pKa value) of ligands used in isostructural pillared

MOFs [Reprinted with permission from ref. 73 (Copyright 2013, Royal Society of Chemistry)]. MOF

Ligand

pKa

DMOF

BDC, DABCO

3.73, 8.86

MOF-508

BDC, BPY

3.73, 4.60

DMOF-TM

TMBDC, DABCO

3.80, 8.86

MOF-508-TM

TMBDC, BPY

3.80, 4.60

97

Table 3 – Characteristics, calculated energy of activation for ligand displacement by water vapor

( ∆E disp ), and experimentally determined maximal structural integrity of selected MOFs (Me3BTC = trimethyl 1,3,5-benzenetricarboxylate, dhBDC = dihydroxybenzenedicarboxylate, MeIm = methylimidizolate, N.A. = not available) [Reprinted with permission from ref. 52 (Copyright 2009, American Chemical Society)].

MOF

Linker

SBU

SBU detail

∆E disp

Experimental

(kcal mol-1)

summary

two 4- and one 6MOF-69C

BDC

chain

coordinate Zn

Unstable

0% steam, ambient

MOF-5

BDC

Zn4O6+

Zn tetramer

11.6

2% steam, 40 °C

MOF-508B

BDC and BPY

Zn22+

Zn paddlewheel

18.9

5% steam, 100 °C

MIL-110

Me3BTC

Al8(OH)159+

Al-octamer

21.2

50% steam, 300 °C

HKUST-1

BTC

Cu22+

Cu paddlewheel

28.9

50% steam, 200 °C

corner-shared Cr-MIL-53

BDC

chain

octahedral

30.5

N.A.

MIL-101

BDC

Cr3OF6+

Cr trimer

35.8

50% steam, 325 °C

MOF-74

dhBDC

chain

edge-shared octahedral

42.0

50% steam, 325 °C

corner-shared

25% steam, 350 °C or

Al-MIL-53

BDC

chain

octahedral

43.4

50% steam, 225 °C

ZIF-8

MeIm

Zn2+

tetrahedral zinc ions

58.5

50% steam, >350 °C

98

Table 4 – Changes in surface areas and pore volumes of unmodified and methyl-modified MIL-

125 samples after immersion in water for different time periods (S = surface area, V = pore volume, m- = methyl-modified) [Reprinted with permission from ref. 157 (Copyright 2014, Royal Society of Chemistry)]. SBET

SLangmuir

Vtotal

Vmicro

Sample

(m2g-1)

(m2g-1)

(cm3g-1)

(cm3g-1)

MIL-125 (as-synthesized)

1550

1700

0.74

0.55

MIL-125 (1 hour)

950

1190

0.83

0.29

MIL-125 (2 hours)

280

440

0.37

0.07

MIL-125 (4 hours)

140

420

0.30

–

m-MIL-125 (as-synthesized)

830

940

0.40

0.30

m-MIL-125 (1 hour)

660

880

0.42

0.26

m-MIL-125 (2 hours)

550

690

0.41

0.18

m-MIL-125 (4 hours)

490

680

0.39

0.15

99

Table 5 – Comparison of properties of pillared MOFs, before and after exposure to 90% RH

[Reprinted with permission from ref. 73 (Copyright 2013, Royal Society of Chemistry)]. Surface area (m2 g-1) Material

Pore volume (cm3 g-1)

Before

After

DMOF

0.75

1980

7

MOF-508

0.42

800

800

DMOF-TM

0.51

1050

1050

MOF-508-TM

0.56

1330

4

Table 6 – List of hydrostable MOFs reported in the article with the respective aqueous

media for which they have been experimentally shown to be moisture-resistant (RTW = room temperature water, BW = boiling water, AM = ambient moisture, S = steam, AAS = aqueous acidic solutions, ABS = aqueous basic solutions). Reported Medium/Media in Hydrostable MOF

which MOF is hydrostable

Reference

(AM, RTW, BW, S, AAS, ABS)

Banasorb-22

AM, RTW, BW, and S

[39]

CH3-MOF-5 and diCH3-MOF-5

AM (up to 4 days)

[42]

MIL-100(Fe)

AM, RTW, and BW

[44]

MOF-177

AM (< 3 days)

[51]

MOF-508B

5% S (100oC)

[52] 100

MIL-110(Al)

50% S (300oC)

[52]

HKUST-1

50% S (200oC)

[52]

MIL-101(Cr)

50% S (325oC)

[52]

MOF-74(Zn)

50% S (325oC)

[52]

MIL-53(Al)

25% S (350oC) or 50% S (225oC)

[52]

ZIF-8

50% S (> 350oC)

[52]

MOF-69C

AM

[52]

UiO-66(Zr)

AM, RTW, and BW

[58]

SCUTC-18

AM and RTW

[60]

SCUTC-19

AM and RTW

[60]

MOF-508

AM and RTW

[73]

DMOF

AM

[73]

DMOF-TM

AM and RTW

[73]

MOF-508-TM

AM

[73]

MIL-125(Ti)

AM, RTW, and BW

[80]

MOF-74(Mg)

AM

[89]

UiO-66-NH2

AM, RTW, and AAS

[91]

UiO-66-NO2

AM, RTW, AAS, and ABS

[91]

UiO-66-Br

AM, RTW, and AAS

[91]

Depends on chemical functionality UiO-67(Zr)

of linker

[93,117]

PCN-222(Fe)

AM, RTW, BW, and AAS

[127]

Cu2TPTC-OR

AM

[133] 101

CALF-25

AM (up to 80oC)

[135]

Nd-MOF

BW

[147]

UiO-66-(CH3)2

AM, RTW, AAS, and ABS

[155]

CH3-MIL-125

RTW (up to 4 hours)

[157]

MOFMC (CNT@MOF-5)

AM (up to 1 week)

[174]

Ni13-MOF-5 and Ni22-MOF-5

AM (up to 1 week)

[178]

MOF-5@SBA-15

AM (up to 1 week)

[181]

H2N-MIL-125-Ti

AM, RTW, BW, and S

[191]

MIL-101(Cr)-NH2

AM, RTW, BW, and S

[192]

102

(b)

(b)

(a) Factors controlling the structural stability of MOFs in aqueous media, and (b) Methods used to improve the hydrostability and hydrothermal cyclic stability of MOFs (PECVD = Plasma Enhanced Chemical Vapor Deposition).

103

•

There are six major factors which control the structural stability of metal organic frameworks (MOFs) in aqueous media.

•

There are eight different strategies which can be adopted to improve the hydrothermal stability of MOFs in aqueous media.

•

Hydrothermal cyclic stability becomes critical when MOFs are being used as adsorbents in adsorption refrigeration systems.

•

Ligand functionalization has been found to be the most effective strategy for improving hydrothermal cyclic stability.

104