1 Introduction
Stability of a coordination polymer (CP) (called also the metal organic framework—MOF) in water is an important subject due to wide range of CPs application in adsorption, (Han et al.
2013) catalysis, (Mounfield III et al.
2016a,
b) as biocomposites, (Doonan et al.
2017) membranes, (Li et al.
2015; Wang et al.
2017) and delivery systems (Cychosz and Matzger
2010). In the case of drug delivery systems the instability of a CP can be utilized during the pH—responsible delivery process (Lin et al.
2016). More or less advanced methods have been applied to determine the stability of CPs in water (Lawrence et al.
2016; Guo et al.
2015; Todaro et al.
2016) however, no strictly defined and widely accepted procedure exists. Recently Gelfand and Shimizu (
2016) proposed six stages of a MOF exposure for the study of stability in water. In their approach the material is subjected to different conditions starting from the near ambient atmosphere (20 °C and 20% of relative humidity) and finishing at the most harshness conditions, i.e., in the boiling water. In contrast, Horwath et al. (
2017) proposed a simpler—one step approach for the stability testing, namely dissolution of 20 mg of a CP in 10 ml of water and a CP filtration after 12 h. One should assure that the process of degradation is in fact a hydrolysis. Thus it should be remembered about the possibility of a CP interaction with different compounds present in a solution—for example with a buffer components (to avoid for example, a precipitation of hardly soluble salts). Also the possible influence of oxygen in the atmosphere during stability testing should be considered, as it was shown by theoretical results of Zhang et al. (
2017). The authors concluded that mixed oxygen/water atmosphere leads to a stronger decrease in a MOF surface area than it was observed for separated oxygen and water atmospheres. This is probably caused by an elongation of the bonds in a CP by the interactions with a gas mixture.
Bezverkhyy et al. (
2016) proposed similar approach to the final stage of Gelfand and Shimizu procedure, and studied the stability of the MIL CPs in water using so called “reflux” method. In this method 100 mg of a CP is suspended in 100 ml of water in an Erlenmeyer flask equipped with a condenser. The solution is heated. Observed decrease in porosity (and the specific surface area) after reflux of fluoride—free CPs in water, was attributed to the blocking of pores by the products of a partial degradation. What is the most important, the XRD data showed no change after the heating in water. The authors concluded that the lack of the changes in the XRD pattern cannot be treated as the proof for the structure stability. Li et al. (
2017) reported recently the details of some CPs degradation process in a phosphate buffer explaining the results reported by Bezverkhyy et al. (
2016).
Also some methods for testing a CP stability in the atmosphere of the vapours of acidic gases have been proposed. For example Han et al. (
2013) first exposed the samples to water and next to SO
2 and NO
2. Mounfield III et al. (
2016a,
b) described the mechanism of degradation of a series of cerium and titanium CPs under the influence of water and acidic gases. The role of defected sites located at the edges of particles was shown to be crucial during the decomposition. The process starts by the acidic gas adsorption on the uncoordinated sites and next it is propagated through the particle. Pang et al. (
2016) pointed out the importance of crystallographic facets stability in the ZIF-8 degradation process at the acidic conditions. The mechanism of the process is based on the water adsorption on Zn(II) and the imidazol linker removal. It was shown that this is easier in the case of (110) facets due to lack of a steric hindrance. The degradation process is additionally assisted by H
2SO
3. Also the susceptibility to transformation of the CPs in water can be crucial for stability, as it was shown for two MONT CPs by Taddei et al. (
2013).
Considering the general rules of a CP degradation in water molecular simulation data are very helpful. De Toni et al. (
2012) studied the initial stage of IRMOF (a hydrophobic material) decomposition using the MD simulation method. It was shown that the process starts by water adsorption on the metallic cluster and next the linker displacement occurs. This leads to the decrease in the CP hydration energy barrier. This mechanism was confirmed by experimental data obtained for different CPs (Pang et al.
2016; Al-Janabi et al.
2016). Thus, for example Todaro et al. (
2016) using the combined approach (including the EPR studies), established the major stages of a HKUST material degradation due to long time exposure to moisture. At the first stage the reversible water adsorption takes place on the majority (65%) of Cu–Cu couples. Next, the remaining 35% of couples adsorb water however, this process is irreversible and leads to the hydrolysis of Cu-O bonds. Since the majority of Cu
2+ ions remains intact, the HKUST is a very water-resistant material. Zuluaga et al. (
2016) studied the process of MOF-74 degradation in water. They showed that hydroxyl created by water splitting is adsorbed on a metal (Zn
2+, Mg
2+, Ni
2+, Co
2+) cation. The process of water splitting is responsible for the elongation of the metal-linker bond, and this effect depends on the type of a metal cation. Tan et al. (
2014) studied the mechanism of water bonding to the MOF-74. It was shown that water is bonded via covalent bonds to the metal cation and via the van der Waals bonds to the linker. Next water dissociation occurs and the dissociation products saturate the metal cation. Interaction of water and Zn from the MOF-5 nodes is also crucial for this MOF stability (Ming et al.
2015) and this stability strongly depends on the relative humidity. Li et al. (
2015) discussed water stability of MOFs applied for the CO
2/N
2 separation. This is important subject since the flue gas contains 5–7% of water. The authors pointed out that the metal–ligand bond strength is crucial and determines the behaviour of a CP during a ligand displacement by water and a CP hydrolysis. Thus, for example usually carboxylic-metal bonds are week and are easily broken by water molecules. The authors also concluded that Cr–O, Mg–O, Al–O and Zn–O bonds have large stability. It was also concluded that to increase a CP stability one should apply the cations having as high oxidation degree as possible. It was also shown that the basicity of the ligand is important for the stability of a CP. Jiao et al. (
2015) discussed the influence of a metal node on the stability of the MOF-74. It was concluded that possibly the standard reduction potential value (larger for Ni
2+ and Co
2+ than for Mg
2+) leading to weaker reducing properties of Ni
2+ than Mg
2+ causes larger stability of Ni-MOF-74 than Mg-MOF-74. Thus it was also concluded that the application of more inert metal as a node leads to larger stability of many other MOFs. Some results (Jiao et al.
2015) confirm this conclusions because the stability of Mg-MOF-74 increases after introduction of Ni
2+ and Co
2+.
However, other simulation data lead to more complicated mechanism of a CP degradation than simple water adsorption and a linker displacement. Bellarosa et al. (
2012) discussed the water resistance of Zn
2+, Be
2+ and Mg
2+ IRMOF-1 structures using the MD approach. It was concluded that the resistance to water is a combination of different CP properties and depends on the core rigidity, a CP flexibility and the strength of coordination of a metal centre by oxygen. Thus, stability depends not only on the type of a metal node but also on the linker (DeCoste et al.
2013). At acidic conditions (HCl) the carboxylate groups can be protonated and some bonds between double rings in linkers can be broken (DeCoste et al.
2013). Tan et al. (
2015) presented very important review and in house experiment on the state of water in a series of isostructural CPs. The stability of CPs with saturated metal centres was shown to be dependent on the strength of metal–oxygen bond and the stability of aminacomplexes. The authors also discussed the methods improving a CP stability in water. Piscopo et al. (
2015) showed that the coordination of a metal cation (Zr
4+ in this case) is crucial for the stability and the preparation conditions influencing this coordination of a cation can change a CP stability. The authors used different methods of a CP synthesis and tested the stability in the HCl (at three pH levels). The XRD data showed no influence of HCl, and the reduction of the size of crystals in HCl was the same as in water. Lu et al. (
2014) performed a series of DFT calculations and concluded that not always a CP stability increases with the coordination number of a metal, and the strength of M–L bond is crucial for this stability. This strength can be estimated by the calculation of the
pKa for the protonanted ligand and the charge of a ligand oxygen atom. Howarth et al. reviewed the most important subjects related to CPs chemical, thermal and mechanical stability (Howarth et al.
2016) i.e., resistance of the structure to degradation. The authors pointed out the lack of standard procedures of stability assessment making the comparison of the results of different studies problematic. Among the most important factors determining a CP stability the authors mentioned: a strength of a linker—metal bond, a shielding of the metal cation by linkers, a metal cation oxidation state, a linker hydrophobicity, application of hydrophobic substituents leading to the reduction of a CP porosity and in this way decreasing the size of water clusters inside. Mounfield III et al. (
2016a,
b) also pointed out the role of H
2SO
3 in MIL-125 degradation showing that the introduction of NH
2 groups stabilizes the CP structure. Jasuja et al. (
2012) reported very interesting approach considering the influence of linkers on the stability of isostructural DMOF materials in water. It was concluded that the introduction of nonpolar groups into a CP structure increases stability and in contrast, the incorporation of a polar group leads to stability decrease. Considering reversibility of degradation MD simulation results show that the both processes: the attack of water molecules on a cation in a CP structure and a linker displacement can be reversible (De Toni et al.
2012). Also initial stages of decomposition can be for some CPs reversible (DeCoste et al.
2013), while the latter stages of HKUST-1 degradation are irreversible (Todaro et al.
2016). There are also reports showing that reversibility/irreversibility depends on relative humidity (Ming et al.
2015; Tan et al.
2014).
Taking into account discussed above results one can conclude that the process of CPs decomposition is very complicated, and the general procedure for the study of the stability in water does not exist. To get some regularities describing CP degradation we decided to start a basic study on the CP decomposition using three nonporous materials, possessing the same linker (Cys) and different metal nodes, having the same oxidation degree but different coordination numbers. To do this a new Ca
2+-containing CP was synthesized and studied together with two freshly reported materials possessing Zn
2+ (Ferrer et al.
2014) and Mg
2+ (Wiśniewski et al.
2018) in their structures. Since all the three materials are nonporous, the next aim of our study is to check the procedure proposed by Gelfand and Shimizu (
2016) for the study of the stability of nonporous CPs. Since the original procedure was proposed for porous CPs we show below, that for nonporous solids the levels of harshness of exposure can be different. We also want to find the general rules causing observed stability of MOFs in water allowing to assign the materials to respective stability group. Additionally, we check the applicability of the XRD and ATR methods for the CP stability testing.
4 Conclusions
The method of the preparation of a new Ca(Cys)2 CP is reported. The structure of this CP has been also determined as Ca(Cys)2·H2O with the CN of the Ca2+ as equal to 7. The stability of studied CPs decreases in the order: Zn(Cys)2 > Mg(Cys)2 > Ca(Cys)2·H2O. It means that the CN of a metal cation is not a decisive factor influencing the stability of studied CPs in water. However, the presence of water molecule in the first coordination shell, observed for Ca(Cys)2·H2O, can be the factor decreasing this CP stability.
The methods used for the study of CPs stability should have been complementary, like XRD and ATR-FTIR. The results for studied nonporous CPs show, that the harshness of exposure is different than for porous CPs. Thus, our results show that immersion in water at 20 °C (ST-4 of the original procedure) is more drastic for studied CPs than the harsh humid conditions (ST-5 the original Gelfand and Shimizu procedure). Therefore the order of both stages should be exchanged and we propose the following arrangement: ST-1: 20 °C–20% R.H., ST-2: 25 °C–50% R.H., ST-3: 50 °C–50% R.H., ST-4: 80 °C–90% R.H., ST-5: 20 °C-immersion in H2O, ST-6: 100 °C-immersion in H2O. Moreover it is proved that 24 h treatment in the case of nonporous CPs is insufficient to test their stability. We propose at least 96 h period of testing. The results of kinetic measurements (ST-5) correlate well with the standard reduction potentials leading to the conclusion that the application of more inert metal as a node leads to larger stability of studied CPs.