Hidden Transformations of a Crystalline Sponge: Elucidating the

Subscriber access provided by University of Pennsylvania Libraries

Article

The Hidden Transformations of a Crystalline Sponge: Elucidating the Stability of a Highly Porous Three-Dimensional MOF Gabriel Brunet, Damir A. Safin, Ilia Korobkov, Andrea Cognigni, and Muralee Murugesu Cryst. Growth Des., Just Accepted Manuscript • DOI: 10.1021/acs.cgd.6b00570 • Publication Date (Web): 17 May 2016 Downloaded from http://pubs.acs.org on May 18, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Crystal Growth & Design is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

The Hidden Transformations of a Crystalline Sponge: Elucidating the Stability of a Highly Porous Three-Dimensional MOF Gabriel Brunet,† Damir A. Safin,† Ilia Korobkov,† Andrea Cognigni,‡ and Muralee Murugesu*,† †

Department of Chemistry and Biomolecular Sciences, University of Ottawa, 10 Marie Curie Private, Ottawa, ON, K1N 6N5, Canada ‡ Department of Chemical Engineering, Norwegian University of Science and Technology, 7491 Trondheim, Norway ABSTRACT: We present our investigations on the stability of a cobalt thiocyanate-based metal-organic framework (MOF), coordinated by the 2,4,6-tris(4-pyridyl)-1,3,5-triazine (TPT) ligand, which is commonly employed for single-crystal-to-single-crystal guest inclusion. The selected crystalline solid, {[(Co(NCS)2)3(TPT)4]·x(solvent)}n, is one emerging MOF currently being utilized as a host for the crystalline sponge method. Our experimental results shed light on two critical structural transformations that should be considered when attempting the revolutionary crystalline sponge method. We demonstrate that when the crystalline sponge is removed from the solution, a remarkable color change from orange to green is observed, yielding a semi-amorphous material. Diffuse reflectance and X-ray absorption spectroscopy studies were carried out on this resulting green compound in order to examine the nature of this transformation. Notably, our analyses indicate a change in the coordination environment of the CoII metal centers from octahedral to tetrahedral at the surface of the material. We further report a second structural transformation of the crystalline sponge, which arises from an increase in the local concentration of Co(NCS)2, and leads to a significant change in the composition and structure of the material, giving a 2D planar sheet network.

1. INTRODUCTION For decades porous materials have been used to trap and store guest substrates with a wide range of potential applications.1−4 Naturally occurring porous materials, such as clays and zeolites, exhibit a variety of pore sizes in the same sample, allowing a lesser degree of control and selectivity for the capture of desired guest molecules.5,6 Hence, researchers have been interested in the smart design of uniform porous materials, by controlling the size, shape and functionalization of the pores.7−13 In this regard, metal-organic frameworks (MOFs) have been shown to be easily tuned and amenable to a wide variety of guest compounds, through well-defined pore spaces that exhibit targeted molecular recognition properties.14−16 Intrinsically, MOFs are well suited to the study of guest inclusion behaviors, due to their potential in exhibiting a singlecrystalline nature, which allows for the structural characterization of the framework with a high degree of accuracy.17,18 Of particular interest are the locations of the preferred sites of adsorption and resulting host-guest bonding motifs. This structural information is essential in the rational improvement of MOF systems for the inclusion of gases and other guests,19−23 however it can be exceedingly difficult to obtain due to the challenges in maintaining adequate crystallinity following guest encapsulation. Among the numerous organic linkers investigated for achieving extended networks, the trigonal 3-connector 2,4,6-tris(4pyridyl)-1,3,5-triazine (TPT) ligand has afforded a large number of crystalline MOFs.24−31 Notably, Fujita and co-workers have successfully demonstrated the remarkable properties of two TPT-based MOFs, namely {[(Co(NCS)2)3(TPT)4]·25(o-

C6H4Cl2)·5(MeOH)}n (1a) and {[(ZnI2)3(TPT)]2·x(solvent)}n, for the capture of different substrates from solution.32,33 This elegant work was centered around the fact that these porous solids can be used as crystalline sponges for performing single-crystal X-ray diffraction (SCXRD) of guest molecules after their efficient capture by the MOF.34,35 This revolutionary approach enables the structural determination of molecules that were previously impossible to crystallize, and in recent years, has been successfully employed by others.36−38 Moreover, transient species or structural intermediates may also be elucidated by the technique. In fact, Duplan et al. have recently confirmed the Michael addition reaction between thiols and cyanoenones as the mechanism-of-action for a new family of covalent drugs.39 Following the success of the technique, Carmalt and co-workers subsequently described the uptake behavior of a series of aromatic molecules within a crystalline sponge, where they adeptly demonstrated the preferential arrangement of guest molecules within the host network, by evaluating the interactions specific to guest functionality.40 Inspired by these outstanding findings, we were intrigued by the possibility of applying the crystalline sponge method for the soaking of magnetic molecules, thus offering a unique opportunity to carry out magneto-structural correlations.41 We initially decided to test the crystalline sponge method using MOF 1a, however, during numerous attempts to prepare the parent MOF, we were repeatedly faced with unexpected and striking experimental results, probably influenced by the observed instability of the reported MOF.34 In order to exploit the full potential of this MOF as a crystalline sponge, it is critical to understand its full physical behavior and stability. With this in mind, the work presented herein provides a de-

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

tailed investigation of the structural transformations displayed by 1a.34 We also offer new insights on the established structure and composition of the porous MOF.

2. EXPERIMENTAL SECTION 2.1. General Considerations. All manipulations, unless otherwise stated, were performed under aerobic/ambient conditions. All materials were used as received from TCI, Strem Chemicals and Sigma Aldrich without further purification. 2.2. Synthesis. 2,4,6-Tris(4-pyridyl)-1,3,5-triazine (TPT)42 and {[(Co(NCS)2)3(κ3-TPT)4]·a(H2O)·b(MeOH)}n (1b)34 were synthesized according to previously reported methods. The removal of single crystals of 1b from the mother liquor, and exposure to aerobic or inert atmospheres, led to the formation of the dark green semi-amorphous material {[(Co(NCS)2)3(κ03 TPT)4]·c(H2O)}n (2). Moreover, extensive evaporation of methanol over a period of one week during the synthesis of 1b led to a change in color of a few crystals from orange to pink resulting in the formation of {[(Co(NCS)2(H2O)0.65(MeOH)0.35)3(κ3-TPT)2]·2.4(H2O)}n (3) and a large amount of microcrystalline pale pink powder, which was also identified as 3. The use of chlorobenzene instead of o-dichlorobenzene during the synthesis of 1b leads to the formation of the pale pink powder of 3 as a major product and very few crystals of 1b. 2.3. Physical Measurements. FTIR spectra were recorded with a Varian 640 FTIR spectrometer equipped with an ATR in the 500–4000 cm–1 range. Diffuse reflectance spectra were measured with a Varian Cary-100 spectrophotometer using polytetrafluoroethylene (PTFE) as a reference. Kubelka-Munk spectra were normalized to allow meaningful comparisons. Thermogravimetric analysis (TGA) data were recorded using a Q5000 IR TGA instrument at a heating rate of 10 °C/min between room temperature and 800 °C, under a constant flow of air or nitrogen (100 mL/min). Powder X-ray diffraction (PXRD) for bulk samples were carried out using a Rigaku Ultima IV X-ray powder diffractometer. The Parallel Beam mode was employed to collect the data (λ = 1.541836 Å). Xray absorption spectra (XAS) were obtained at the SwissNorwegian Beamline (SNBL) at the European Synchrotron Radiation Facility (ESRF) in Grenoble, France. The data was collected at room temperature on the BM01B setup. The measurements were carried out using the transmission mode between 7.65 and 8.50 keV, with a step of 0.4 eV, and averaged. 2.4. Single Crystal X-ray Diffraction Studies. Data collection results for 1b and 3, represent the best data sets obtained in several trials for each sample. Prior to data collection, crystals of 1b were preserved in the mother liquor and subsequently mounted on thin glass fibers using paraffin oil. The data set for the crystal of 3 was collected at room temperature using colorless nail polish to attach the crystal to the thin glass fiber pin. The data was collected on a Bruker AXS KAPPA single crystal diffractometer equipped with a sealed Mo tube source (λ = 0.71073 Å) APEX II CCD detector. Raw data collection and processing were performed with APEX II software package from Bruker AXS.43 Diffraction data for both samples were collected with a sequence of 0.3° ω scans at 0, 90, 180 and 270° in φ to provide the acceptable redundancy of data. Initial unit cell parameters were determined from 60 data frames with 0.3° ω scan each collected at the different sections of the Ewald sphere. Semi-empirical absorption corrections based on equivalent reflections were applied.44 Systematic

Page 2 of 9

absences in the diffraction data-set and unit-cell parameters were consistent with trigonal R–3m (№ 166) space group for 1b and trigonal R–3 (№ 148) space group for 3. Solutions in the centrosymmetric space groups for both compounds yielded chemically reasonable and computationally stable results of refinement. The structures were solved by direct methods, completed with difference Fourier synthesis, and refined with full-matrix least-squares procedures based on F2. All hydrogen atoms were treated as idealized contributions during the refinement. All scattering factors are contained in several versions of the SHELXTL program library, with the latest version used being v.6.12.45 Crystallographic data and selected data parameters are reported in Table 1. The crystal structures have been deposited at the Cambridge Crystallographic Data Centre and allocated the deposition numbers CCDC 1008410 (1b) and 1008411 (3). Table 1. Summary of the crystal structure data and refinement for compounds 1a, 1b and 3. 1a34 1b 3 C26H16N10CoS2 C78H48Co3N30S6 C44.10H45Co3N18 O8.40S6 FW, g mol-1 591.56 1774.61 1330.73 crystal syscubic trigonal trigonal tem T, K 90(2) 200(2) 296(2) a, Å 37.461(5) 26.7888(6) 26.6894(6) b, Å 37.461(5) 26.7888(6) 26.6894(6) c, Å 37.461(5) 65.6133(18) 9.8119(3) α, o 90 90 90 β, o 90 90 90 γ, o 90 120 120 V, Å3 52 571(12) 40 778(2) 6052.9(3) Z 24 6 3 ρcalcd, g cm-3 0.448 0.434 1.095 µ (Mo, Kα), 0.255 0.246 0.812 mm-1 reflns col1954 5084 3286 lected R1, wR2 (I > 0.0998, 0.2830 0.0600, 0.1650 0.0670, 0.1867 2 σ (I))a R1, wR2 (all 0.1286, 0.2978 0.1259, 0.2095 0.0878, 0.2087 data) a R1 = (||Fo| – |Fc||)/|Fo|; wR2 = {[w(Fo2 – Fc2)2]/[w(Fo2)2]}1/2, where w = 1/[σ2(Fo2) + (aP)2 + bp], where P = [max(0, Fo2) + 2Fc2]/3; and Rw = [w(|Fo| – |Fc|)2/w|Fo|2]1/2, where w = 1/σ2(|Fo|). Complex Formula

3. RESULTS AND DISCUSSION Synthesis. Crystalline sponges have tremendous potential applications in host-guest chemistry and modern X-ray crystallography due to their ability to elucidate crystal structures of non-crystallizing systems. Our numerous attempts to obtain the reported compound 1a by carefully reproducing the synthetic procedure were unsuccessful,34 and continuously led to the formation of 1b. While the connectivity of the porous MOF remains the same in 1a and 1b, we have determined that the correct lattice solvents consist of water and/or methanol, as evidenced by SCXRD and other spectroscopic analyses (vide infra). The initial report of 1a suggested that the pores were filled with a significant amount of o-dichlorobenzene, which


Page 3 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


was not readily apparent in the X-ray structure or other analyses. Furthermore, during our manipulations, we have observed a progressive and gradual color change in the single crystals of 1b, from light orange to green, when they are removed from solution and kept under aerobic or inert atmospheres. This results in the formation of a dark green semi-amorphous material (2) (Figure 1 and Figure S1 for PXRD in the Supporting Information). During this process, visible cracking of the crystals, as well as a slight volume contraction, can be observed. To monitor this transformation, single crystals of 1b were removed from the mother liquor and their time evolution at ambient conditions was followed by optical microscopy (Figure 2). Interestingly, the presence of heat, e.g. from a light source, accelerates this irreversible process. To date, such a transformation of the crystalline sponge has yet to be reported. This critical irreversible transformation to a semi-amorphous material prevents any further use of this system as a crystalline sponge for SCXRD applications due to the partial loss of crystallinity and structural change. Therefore, in order to maintain the structural integrity of 1b out of solution, the compound must be immediately immersed in oil. Alternatively, single-crystals of 1b may be safely stored in noncoordinating solvents such toluene and o-dichlorobenzene, preventing any structural transformations.

Figure 1. Synthesis of 1b−3 and their corresponding crystal pictures as viewed under an optical microscope at room temperature.

Figure 2. Photos highlighting the gradual transformation of three crystals of 1b to 2 over time, viewed under optical microscope at ambient conditions.

Furthermore, if 1b is left in its mother liquor over a period of one week, a change in color of the crystals, from orange to pink, can be observed along with the formation of a microcrystalline pink powder. This transformation results in the new compound, 3. The co-crystallizing pink powder was also

identified as 3 (Figure S1). This notable structural transformation can be explained by the evaporation of the layered methanol, resulting in a local increase of the Co(NCS)2 concentration. Moreover, the replacement of o-dichlorobenzene with chlorobenzene during the synthesis of 1b leads to the direct formation of 3 in the form of pale pink powder as the major product and few crystals of 1b. This remarkable difference can be explained by the significantly lower density and viscosity of chlorobenzene compared to o-dichlorobenzene, which, in turn, supports faster diffusion of the methanol layer of Co(NCS)2 into the chlorobenzene solution of TPT. Subsequently, the reaction rate between Co(NCS)2 and TPT increases, favoring the formation of 3. Structural analysis. Crystals of 1b and 3 suitable for SCXRD analysis were obtained from mother liquors during the synthesis. It should be noted that several different crystals of 1b and 3 have been checked by SCXRD, testifying to their identity. Recently, the crystal structure of 1a was reported to be refined in the cubic space group Fm–3m.34 It is well known that several criteria have to be considered for a correctly refined structure. Among them, SHELXL calculates the most important criteria, namely R1 (observed data, I > 2σ(I)), wR2 (all data) and the so-called goodness of fit, S. Generally acceptable values for these criteria are 2σ(I)) = 6.00%, wR2(all) = 20.95%, S = 1.04, support the suggested model. Although there is a strong disagreement in the refinement of the X-ray data, the overall crystal structure of 1b closely resembles the structure of 1a, and therefore, only the main features will be examined. In particular, the 3D coordination network reveals an infinite aggregation of three cage-like frameworks, viz. octahedral [Co6(NCS)12(κ3-TPT)4], cuboctahedral [Co12(NCS)24(κ3-TPT)8] and [Co12(NCS)24(κ3-TPT)24] polyhedra. The topology of the network was analyzed by the program package TOPOS.46 The underlying net of 1b can be described as a (3-coordinated)4(4-coordinated)3 2-nodal network with Schläfli symbol {62.82.102}3{63}4 formed by 3connected TPT ligands and molecular fragments of [Co(NCS)2] as nodes. Hence, this network is identified by the twisted boracite (tbo) net found in the RCSR database. The coordination environment surrounding the cobalt metal centers is comprised of four nitrogen atoms originating from the pyridine moieties of four TPT ligands and two terminally bonded thiocyanate groups, forming an octahedral coordination environment (Figure 3). Notably, there are two crystallographically independent cobalt atoms (Figure S2), with Co1 displaying a 4-fold rotational symmetry axis along the axial positions of the compressed octahedron, while Co2 is in a lower symmetry environment with a slightly distorted octahedral geometry (Table S1). The NCS– anions were found to be N-bound to CoII in a pronounced linear fashion with the (Co)N–C–S and Co–N–C(S) bond angles being almost 180° (Table S1). The Co–N(Py) bond lengths are similar with values ranging from 2.17 to 2.21 Å, while the Co–N(thiocyanate) distances are



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

significantly shortened with values of about 2.07 Å (Table S1). The shortest Co···Co distances between adjacent and

Page 4 of 9

opposite metal centers in the [Co6(NCS)12(κ3-TPT)4]

Figure 3. Crystal structures of three discrete molecular cages contained within 1b and spacefill representation of the octahedral

[Co6(NCS)12(κ3-TPT)4] cage. Color code: purple (Co), blue (N), gray (C), yellow (S). Hydrogen atoms were omitted for clarity. polyhedron have been determined to be ∼13.40 and ∼18.94 Å. Furthermore, the X-ray analysis revealed that all pyridine rings of the TPT ligand were 50% disordered over two overlapping positions. The volume of the potential void in 1b was calculated to be 30334.2 Å3, which is about 74.4% of the unit cell volume. Such large voids were proven to be critical in encapsulating guest molecules,34 thus making 1b a desirable crystalline sponge. In order to maintain this attractive feature, the structural transformation of 1b to 2 must be avoided. The instability of the crystalline sponge, during both manipulations and synthesis, render 1b a significantly more difficult challenging host to carry out single-crystal-to-single-crystal transformations. The zinc-based crystalline sponge,35 which exhibits a higher stability, is therefore recommended. These results highlight the growing demand for crystalline sponges with pore sizes suitable for larger guests. In our efforts to identify the solvent molecules contained within the pores of 1b, we were surprised to repeatedly observe the absence of o-dichlorobenzene molecules by SCXRD. After completing a chemically reasonable and computationally stable model for the core of the structure, Fourier maps difference revealed several well defined residual electron density peaks scattered in the void space of the structure. From all the possible solvent species involved in the synthesis, it was a challenge to confirm the presence of o-dichlorobenzene in the cavities of the MOFs. In order to evaluate this, the strength of the most prominent residual electron density peaks should be

evaluated (see SI for details). Considering the quality of the data set, it would be highly speculative to draw conclusions from the positions of the peaks with such low overall intensity. Nevertheless, it should be noted that the characteristic shapes of aromatic rings were absent in the voids of the MOF. Hence, unit cell formulation cannot be carried out exclusively from the results of the X-ray analysis and would require additional experimental confirmation from complementary analytical experiments using the bulk of the compound. In contrast to the observed 3D structure of 1b, the molecular structure of compound 3, refined in the trigonal space group R–3, exhibits tightly packed 2D sheets that are rotated 60° relative to each other, with an interlayer distance between two adjacent sheets of ∼3.25 Å (Figure 4). It was found that this interlayer region is partly filled with lattice water molecules and further stabilized by π-π stacking interactions between the triazine rings of the TPT ligand which are aligned in a parallel fashion. Furthermore, the sulfur atom of each NCS− anion participates in hydrogen bonding with a hydrogen atom of a coordinated water molecule belonging to an adjacent sheet and also interacts with the π system of other NCS− anions that are close in proximity. Subsequently, we observe the formation of 2D sheets, that are constructed from fragments of [Co(NCS)2(H2O)0.65(MeOH)0.35] and linked by the TPT ligand. The underlying net of 3 was also determined considering each TPT ligand as a three-connected linker and each


Page 5 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design the Schläfli symbol {63}. Therefore, 3 can be identified by the hcb (Shubnikov hexagonal plane net) topology. All CoII atoms

[Co(NCS)2(H2O)0.65(MeOH)0.35] fragment as a node. The resulting topology is a three-connected uninodal network with

3

Figure 4. A) Partially labeled molecular structure of the [Co(NCS)2(κ -TPT)2(H2O)2] fragment of 3. B) Crystal packing of 3 along the caxis, with different colors (pink, cyan and orange) corresponding to a different layer. The coordination polyhedra display the location of the cobalt metal centers with respect to their sheet. C) Crystal packing of 3 along the b-axis, displaying the planar sheet arrangement. Color code: purple (Co), blue (N), gray (C), yellow (S), red (O). Hydrogen atoms, MeOH and lattice H2O molecules were omitted for clarity.

are six coordinate and adopt a slightly distorted octahedral geometry, formed by two nitrogen atoms of two distinct TPT pyridyl groups (N2, N2a), two nitrogen atoms of the NCS– anions (N3, N3a) and two oxygen atoms of the water and/or methanol molecules (O1, O1a). The latter solvent molecules exhibit partial occupancies of 65 and 35%, respectively. Moreover, the NCS– anions were found to be N-bound to CoII in a slightly bent fashion with the (Co)N–C–S and Co–N–C(S) bond angles being about 178 and 172°, respectively, while the Co–N(Py) and Co–N(thiocyanate) bond lengths have been determined to be 2.15 Å and 2.08 Å, respectively (Table S2). The shortest intralayer Co···Co distance (through the TPT ligand) is 13.35 Å, while the shortest Co···Co distance occurs between adjacent layers with a distance of 8.37 Å. The potential volume of the void space in 3, estimated by removing the lattice water molecules, was calculated to be 1792.4 Å3, which is about 29.6% of the total unit cell volume (Figure S3). Close comparison of the structures of 1b and 3, reveals CoII atom that are coordinated by either four (for 1b) or two (for 3) TPT ligands, indicating an increase of the metal to TPT ratio from 1b to 3. Therefore, an increase of the local concentration of Co(NCS)2 during the reaction with TPT would promote a change in the structural organization of the 3D coordination network towards a tightly packed 2D coordination network. In addition, the observed void volume is nearly three-fold smaller than the one observed in 1b and the accessibility of the void space is limited due to the coordinated NCS− anions and water/methanol molecules, making complex 3 a more challenging host for the crystalline sponge method. In order to provide further insights into the three aforementioned materials, we have performed a series of spectroscopic analyses. Specifically, these studies will prove essential in elucidating the solidstate transformation from 1b to 2 due to its semi-amorphous nature.

IR spectroscopy analysis. The FTIR spectra of 1b−3 each contain a band characteristic of the CN stretches of the NCS– anions, at 2050, 2085 and 2055 cm–1, respectively (Figure S4). In all three cases, these values confirm the coordination of the NCS– anions through the nitrogen atoms with values below 2100 cm-1. The spectrum of 3 exhibits a broad band, centered at about 3350 cm–1, for the symmetric and asymmetric OH stretchings of the H2O and MeOH molecules,47 while the spectrum of 2 shows a similar band with two maxima at 3285 and 3400 cm–1. The latter finding testifies to the presence of H2O molecules in the structure of 2. The same band in the FTIR spectrum of 1b is absent, however, due to a significant broadening. Furthermore, water molecules were observed in the FTIR spectra of 1b−3 as characteristic weak bands for the HOH bending at 1615, 1610 and 1655 cm–1.47 The lowfrequency region of the FTIR spectra of 1b−3 also exhibit a band, characteristic for the so-called librational modes of the H2O molecules,47 at about 745 cm–1. This band is due to the rotational oscillations of the water molecule, restricted by interactions with neighboring atoms. Moreover, the spectrum of 1b displays a set of pronounced bands, characteristic for MeOH molecules, at 1035, 1435 and 1455 cm–1. These bands are significantly less intense in the spectrum of 3, which is obviously due to a lesser amount of MeOH molecules in the structure of 3 compared to that of 1b. These bands were not found in the spectrum of 2, testifying to the absence of MeOH in this complex. Thus, the remarkable solid-state-to-solid-state transformation of 1b to 2 can be explained by the spontaneous evaporation of the MeOH molecules, trapped in the cages of the former complex. This might also be supported by the fact that this transformation can occur in both aerobic and inert atmospheres, and hence, cannot be due to oxidation processes. Additionally, the FTIR spectra of 1b−3 contain all the characteristic bands for the parent TPT ligand (Figure S4).



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Thermogravimetric analysis. In order to determine the amount of solvent contained in the semi-amorphous material 2, TGA experiments were carried out in comparison to that of the parent porous MOF. The TGA analysis data of 1b and 2 indicates that the latter complex resembles the composition of the former one, but exhibits a significantly lesser amount of solvent (Figure S5), following the solid-state-to-solid-state transformation. It should be noted that the TGA curves of 1b and 2 in a dynamic air atmosphere are the same as those in a dynamic nitrogen atmosphere. Consequently, both complexes are completely burned with the formation of a stable residue regardless of the atmosphere, which can be explained by either the presence of oxygen trapped in the porous structure, or the catalytic release of oxygen from solvents (H2O and/or MeOH) during the thermal decomposition. It was found that the molecule of 1b is unstable even at room temperature, and starts to decompose due to an extensive loss of solvent, which is shown in three steps and corresponds to about 70% mass loss. The remaining two decomposition steps correspond to the burning of the TPT and NCS– ligands. Compound 2 is stable up to about 35 °C and decomposes in a very similar fashion as 1b, however, the mass loss due to solvent evaporation in 2 is significantly less than in 1b and constitutes approximately 35% of the total mass. In contrast to the TGA curves observed in 1b and 2, the thermal decomposition of 3 shows only a slight loss of mass up to 200 °C (approximately 10%), owing to the smaller amount of solvent molecules contained in the tightly packed 2D sheets (Figure S5). Diffuse reflectance spectroscopy analysis. In order to further investigate the solid-state-to-solid-state transformation of 1b to 2, the pure solid materials were analyzed by diffuse reflectance spectroscopy at room temperature. The diffuse reflectance spectrum of TPT contains a band with a shoulder exclusively in the UV region, which is attributed to n–π* and π–π* transitions (Figure 5). The diffuse reflectance spectra of 1b, measured over time, each contain intense bands in the UV region arising from the same transitions of the organic ligand (Figure 6, A). These bands are accompanied with low-intense gradual shoulders up to about 540 nm, which correspond to the ligand-to-metal and/or metal-to-ligand charge transfers. The spectrum of the complex at 0 h contains a broad band in the visible region, centered at 760 nm, corresponding to the d– d transitions.

Figure 5. Normalized Kubelka-Munk spectra for TPT and 3.

Page 6 of 9

The observable band is caused by a 4T1g(F)→4T1g(P) transition. It is easy to establish that the 4T1g(F)→4T2g transition should correspond to the value of energy which is slightly lower in comparison with the energy of the 4T1g(F)→4T1g(P) transition. Due to the 4A2g state arising from the t32ge4g configuration and the 4T1g(F) state mainly from t52ge2g, the 4 T1g(F)→4A2g transition is in essence a two-electron transfer, and for this reason, the band should be less intense (approximately by two orders of magnitude), than the bands of other transitions. The two aforementioned circumstances for a 4 T1g(F)→4A2g transition, namely small intensity and affinity to the band 4T1g(F)→4T1g(P), lead to the fact that this transition is not observed in the spectrum. At the same time, the band at 760 nm is accompanied by a shoulder at ∼640 nm, which is believed to originate from spin-orbit effects. The absorption band at 760 nm in the spectrum of 1b at 0 h gradually decreases and exhibits a red-shift (18 nm) over longer intervals of time (Figure 6, B). This process is accompanied with the gradual appearance of a new broad band, with three maxima, centered at 560, 600 and 635 nm (Figure 6, C). The high-energy band corresponds to a transition from the basic 4A2 state to a 4 T1(P) state. The thin structure is caused by spin-orbital interactions as a result of which, first, there is a splitting of the 4 T1(P) state and, secondly, there are resolved transitions in the next doublet states with the same intensity. The former lowenergy band corresponds to the 4A2→4T1(F) transition, while the third possible 4A2→4T2 transition is outside of the visible region. Thus, the diffuse reflectance spectroscopy data unequivocally confirms a tetrahedral environment for the CoII cations in 2 upon the crystalline-amorphous phase transition. This can be rationalized by the evacuation of MeOH molecules from the lattice (vide supra), leading to a structural collapse of the large pores of 1b. Thus, the NCS− ligands would remain coordinated, as evidenced through spectroscopic analyses, while two of the four TPT linkers would break off and rearrange, yielding a tetrahedral coordination environment. Similar changes in the coordination geometry of cobalt metal centers have been previously reported, and are often attributed to atomic rearrangements at different temperature ranges and/or the loss of guest molecules.48−50

Figure 6. Normalized Kubelka-Munk spectra depicting the color change from orange (1b) to green (2) over time, with the black arrows displaying the significant changes.


Page 7 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The kinetics of the solid-state-to-solid-state transformation was studied by examining the Kubelka-Munk function at 760 nm (Figure 7). The time dependent evaluation of these bands can be divided in two steps. During the first step (t < 0.5 h), the F(R) values exhibit a slow decrease, which can be fitted with a first-order exponential equation, allowing the calculation of the kinetic constant kI = 0.0549 h−1. The second step evaluates up to about 12 h and can be also fitted with a firstorder exponential equation yielding a kinetic constant of kII = 0.9436 h–1 (Figure 7). The relatively constant values for F(R) during the first half-hour of this transformation, suggests that crystals of 1b can be stable up to 30 mins when kept out of solution, even when physical changes in the crystals can be observed (Figure 2). This solid-state-to-solid-state transformation, however, appears to be irreversible after 30 mins as evidenced by the drastic change in kinetics.

prevalent in the diffuse reflectance spectroscopy data, while the enol form is exclusively represented in the crystal structure.51−55 In order to verify this postulation, we have performed a second time-dependent diffuse reflectance spectroscopy study, where the sample is crushed following the observation of tetrahedral bands in the spectra (Figure S7). Following 7 h of single crystals of 1b being exposed to ambient conditions, we have observed an identical behavior as previously described, where bands centered at ∼600 and 760 nm are clearly identifiable, indicating the presence of both tetrahedral and octahedral coordination environments, respectively. Following the immediate crushing of the sample at 7 h, we have witnessed a pronounced increase in the intensity of the band corresponding to an octahedral species, which is accompanied by a considerable decrease in the intensity of the tetrahedral band. This behavior can be explained by the formation of a new surface, which is predominantly octahedral, and confirms that the change in coordination environment occurs primarily at the surface of compound 2. Interestingly, the newly formed surface continues to convert the symmetry of the Co ions from octahedral to tetrahedral after 7 h, as evidenced by DRS (Figure S7).

4. CONCLUSION

Figure 7. Time dependence of the normalized Kubelka-Munk function F(R) of 1b at 760 nm and 298 K.

Complex 3 was also analyzed by diffuse reflectance spectroscopy in the solid-state (Figure 5). The spectrum exhibits an intense band, accompanied with a shoulder, in the UV region with the maxima at 220 and 350 nm, respectively. These bands were attributed to intraligand n–π* and π–π* transitions. The ligand-to-metal and/or metal-to-ligand charge transfers were shown as a low intense band from about 450 to 640 nm, while the characteristic band centered at about 760 nm was assigned to the d–d transition, testifying to an octahedral coordination environment around the metal center in 3. X-ray absorption spectroscopy analysis. To complement the diffuse reflectance spectroscopy data and further confirm the coordination environment of 2, we carried out X-ray absorption spectroscopy experiments. To our surprise, the XANES part of the spectra revealed the presence of a small pre-peak before the edge position, suggesting that the main symmetry of the cobalt cations is octahedral (Figure S6). Furthermore, the relative small intensity of this feature is coherent with the fact that the octahedral symmetry is distorted. Accordingly, we suspect that only the exposed surface of 2 undergoes a change in coordination environment to tetrahedral, while the overwhelming majority of the material remains octahedral. Similar phenomena have been previously described in photochromic N-salicylidene aniline derivatives, where the keto form is

In conclusion, we have reported the well-refined X-ray structure of a crystalline sponge that is employed in a revolutionary crystallographic approach.34 Our investigations into the stability of a CoII-based MOF have afforded important insights into two surprising structural transformations. The instability of the porous MOF results in a gradual and observable solid-state transformation towards a semi-amorphous material, where the coordination environment of the metal centers is significantly altered. It should be noted however, that this irreversible transformation occurs at the surface of the single-crystals rather than the bulk of the material. Furthermore, we have revealed that careful monitoring of the reaction vessel during the synthesis of the crystalline sponge is essential to prevent methanol evaporation, which could increase the local concentration of Co(NCS)2 and result in a drastic change in the coordination network, from three-dimensional to two-dimensional sheets. These findings are critical to take into account if further guest inclusion studies involving the Co-based crystalline sponge are to be carried out. We believe that it will assist in the encapsulation of guest molecules through a fundamental understanding of the stability of the crystalline sponge.

ASSOCIATED CONTENT Supporting Information Crystallographic data (including CIF files), TGA curves, FTIR, XAS and additional DRS spectra are included. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author *Email: [email protected]

Notes The authors declare no competing financial interests.

ACKNOWLEDGMENTS We thank the University of Ottawa, the Canadian Foundation for Innovation (CFI), NSERC and ORF for their financial support.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

We also gratefully acknowledge Dr. K. Robeyns and Prof. Y. Filinchuk (Université catholique de Louvain, Louvain-la-Neuve, Belgium) for assistance with single crystal X-ray analysis and Prof. J. Brusso for access to the UV-Vis spectrophotometer. The authors also thank R. J. Holmberg for providing PXRD data and Dr. W. van Beek (Swiss-Norwegian Beamlines (SNBL) at European Synchrotron Research Facility (ESRF), Grenoble, France) for providing XAS data.

REFERENCES (1) Li, J.-R.; Yu, J.; Lu, W.; Sun, L.-B.; Sculley, J.; Balbuena, P. B.; Zhou, H.-C. Nat. Commun. 2013, 4, No. 1538. (2) Fryxell, G. E.; Liu, J.; Hauser, T. A.; Nie, Z.; Ferris, K. F.; Mattigod, S.; Gong, M.; Hallen, R. T. Chem. Mater. 1999, 11, 2148−2154. (3) Seo, J. S.; Whang, D.; Hyoyoung, L.; Jun, S. I.; Oh, J.; Jeon, Y. J.; Kim, K. Nature 2000, 404, 982−986. (4) Davis, M. E. Nature 2002, 417, 813−821. (5) Diamond, S. Clays Clay Miner. 1970, 18, 7−23. (6) Ackley, M. W.; Rege, S. U.; Saxena, H. Microporous Mesoporous Mater. 2003, 61, 25−42. (7) Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O’Keeffe, M.; Yaghi, O. M. Science 2002, 295, 469−472. (8) Banerjee, R.; Furukawa, H.; Britt, D.; Knobler, C.; O’Keeffe, M.; Yaghi, O. M. J. Am. Chem. Soc. 2009, 131, 3875−3877. (9) Farha, O. K.; Malliakas, C. D.; Kanatzidis, M. G.; Hupp, J. T. J. Am. Chem. Soc. 2010, 132, 950−952. (10) Kaye, S. S.; Dailly, A.; Yaghi, O. M.; Long, J. R. J. Am. Chem. Soc. 2007, 129, 14176−14177. (11) Dincă, M.; Dailly, A.; Tsay, C.; Long, J. R. Inorg. Chem. 2008, 47, 11−13. (12) Férey, G.; Mellot-Draznieks, C.; Serre, C.; Millange, F. Acc. Chem. Res. 2005, 38, 217−225. (13) Katsenis, A. D.; Puškarić A.; Štrukil, V.; Mottilo, C; Julien, P. A.; Užarević, K.; Pham, M.-H.; Do, T.-O.; Kimber, S. A. J.; Lazić, P.; Magdysyuk, O.; Dinnebier, R. E.; Halasz, I.; Friščić, T. Nat. Commun. 2015, 6, 6662. (14) Chen, B.; Xiang, S.; Qian, G. Acc. Chem. Res. 2010, 43, 1115−1124. (15) Kreno, L. E.; Leong, K.; Farha, O. M.; Allendorf, M.; Van Duyne, R. P.; Hupp, J. T. Chem. Rev. 2012, 112, 1105−1125. (16) Li, J.-R.; Sculley, J.; Zhou, H.-C. Chem. Rev. 2012, 112, 869−932. (17) Yang, S.; Sun, J.; Ramirez-Cuesta, A. J.; Callear, S. K.; David, W. I. F.; Anderson, D. P.; Newby, R.; Blake, A. J.; Parker, J. E.; Tang, C. C.; Schröder, M. Nat. Chem. 2012, 4, 887−894. (18) Haneda, T.; Kawano, M.; Kojima, T.; Fujita, M. Angew. Chem. Int. Ed. 2007, 46, 6643−6645. (19) Rowsell, J. L. C.; Spencer, E. C.; Eckert, J.; Howard, J. A. K.; Yaghi, O. M. Science 2005, 309, 1350−1354. (20) Yang, C.; Wang, X.; Omary, M. A. Angew. Chem. 2009, 121, 2538−2543. (21) Vaidhyanathan, R.; Iremonger, S. S.; Shimizu, G. K. H.; Boyd, P. G.; Alavi, S.; Woo, T. K. Science 2010, 330, 650−653. (22) Zhou, H.-L.; Lin, R.-B.; He, C.-T.; Zhang, Y.-B.; Feng, N.; Wang, Q.; Deng, F.; Zhang, J.-P.; Chen, X.-M. Nat. Commun. 2013, 4, 2534. (23) Cox, J. M.; Walton, I. M.; Benson, C. A.; Chen, Y.-S.; Benedict, J. B. J. Appl. Cryst. 2015, 48, 578−581. (24) Neville, S. M.; Halder, G. J.; Murray, K. S.; Moubaraki, B.; Kepert, C. J. Aust. J. Chem. 2013, 66, 452−463. (25) Gamez, P.; Reedijk, J. Eur. J. Inorg. Chem. 2006, 29−42. (26) Arcís-Castillo, Z.; Muñoz, M. C.; Molár, G.; Bousseksou, A.; Real, J. Chem. Eur. J. 2013, 19, 6851−6861.

Page 8 of 9

(27) Fu, Z.; Chen, Y.; Zhang, J.; Liao, J. Mater. Chem. 2011, 21, 7895−7897. (28) Chang, Z; Zhang, D.-S.; Hu, T.-L.; Bu, X.-H. Inorg. Chem. Commun. 2011, 14, 1082−1085. (29) Batten, S. R.; Hoskins, B. F.; Robson, R. J. Am. Chem. Soc. 1995, 117, 5385−5386. (30) Batten, S. R.; Harris, A.R.; Murray, K. S.; Smith, J. P. Cryst. Growth Des. 2002, 2, 87−89. (31) Barrios, L. A.; Ribas, J.; Aromí, G.; Ribas-Ariño, J.; Gamez, P.; Roubeau, O.; Teat, S. J. Inorg. Chem. 2007, 46, 7154−7162. (32) Inokuma, Y.; Arai, T.; Fujita, M. Nature Chem. 2010, 2, 780−783. (33) Biradha, K.; Fujita, M. Angew. Chem. Int. Ed. 2002, 41, 3392−3395. (34) Inokuma, Y.; Yoshioka, S.; Ariyoshi, J.; Arai, T.; Hitora, Y.; Takada, K.; Matsunaga, S.; Rissanen, K.; Fujita, M. Nature 2013, 495, 461−466. (35) Ramadhar, T. R.; Zheng, S.-L.; Chen, Y.-S.; Clardy, J. Acta Cryst. 2015, A71, 46−58. (36) Vinogradova, E. V.; Müller, P.; Buchwald, S. L. Angew. Chem. 2014, 126, 3189−3192. (37) Mori, K.; Akasaka, K.; Matsunaga, S. Tetrahedron 2014, 70, 392−401. (38) Ramadhar, T. R.; Zheng, S.-L.; Chen, Y.-S.; Clardy, J. Chem. Commun. 2015, 51, 11252−11255. (39) Duplan, V.; Hoshino, M.; Li, W.; Honda, T.; Fujita, M. Angew. Chem. Int. Ed. 2016, 55, 4919−4923. (40) Hayes, L. M.; Knapp, C. E.; Nathoo, K. Y.; Press, N. J.; Tocher, D. A.; Carmalt, C. J. Cryst. Growth Des. 2016, DOI: 10.1021/acs.cgd.6b00435. (41) Aulakh, D.; Pyser, J. B.; Zhang, X.; Yakovenko, A. A.; Dunbar, K. R.; Wriedt, M. J. Am. Chem. Soc. 2015, 137, 9254−9257. (42) Li, M.-X.; Miao, Z.-X.; Shao, M.; Liang, S.-W.; Zhu, S.-R. Inorg. Chem. 2008, 47, 4481−4489. (43) APEX Software Suite v. 2010, Bruker AXS, Madison, WI, 2005. (44) Blessing, R. Acta Cryst. 1995, A51, 33−38. (45) Sheldrick, G. M. Acta Cryst. 2008, A64, 112−122. (46) Blatov, V. A.; Shevchenko, A. P.; Proserpio, D. M. Cryst. Growth Des. 2014, 14, 3576−3586. (47) Nakamoto, K. Infrared and Raman Spectra of Inorganic and Coordination Compounds. Wiley: N. Y. 1997, Part B, 5th Edition, pp. 116. (48) Chen, C.-L.; Goforth, A. M.; Smith, M. D.; Su, C.-Y.; zur Loye, H.-C. Angew. Chem. Int. Ed. 2005, 44, 6673−6677. (49) Zeng, M.-H.; Hu, S.; Chen, S.; Xie, G.; Shuai, Q.; Gao, S.-L.; Tang, L.-Y. Inorg. Chem. 2009, 48, 7070−7079. (50) Zeng, M.-H.; Tan, Y.-X.; He, Y.-P.; Yin, Y.; Chen, Q.; Kurmoo, M. Inorg. Chem. 2013, 52, 2353−2360. (51) Safin, D. A.; Robeyns, K.; Garcia, Y. CrystEngComm. 2012, 14, 5523−5529. (52) Safin, D. A.; Robeyns, K.; Garcia, Y. RSC Adv. 2012, 2, 11379–11388. (53) Safin, D. A.; Bolte, M.; Garcia, Y. CrystEngComm 2014, 16, 5524–5526. (54) Safin, D. A.; Babashkina, M. G.; Robeyns, K.; Bolte, M.; Garcia, Y. CrystEngComm 2014, 16, 7053–7061. (55) Safin, D. A.; Bolte, M.; Garcia, Y. CrystEngComm 2014, 16, 8786–8793.


Page 9 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


For Table of Contents Use Only The Hidden Transformations of a Crystalline Sponge: Elucidating the Stability of a Highly Porous Three-Dimensional MOF Gabriel Brunet, Damir A. Safin, Ilia Korobkov, Andrea Cognigni and Muralee Murugesu*

Synopsis: The structural stability of a benchmark crystalline sponge is evaluated. We characterize two critical structural transformations that affect not only the crystallinity of the host compound, but also its ability to uptake desired guest molecules.


9

Hidden Transformations of a Crystalline Sponge: Elucidating the

Recommend Documents