Microporous and Mesoporous Materials 360 (2023) 112688 Contents lists available at ScienceDirect Microporous and Mesoporous Materials journal homepage: www.elsevier.com/locate/micromeso Examining the adsorption of gases into solid crystalline molecular copper (II) 3,5-bis(trifluoromethyl)benzoate derivatives Oriol Vallcorba a, Julio Fraile-Sainz b, Concepción Domingo b, José A. Ayllón c, * a ALBA Synchrotron Light Source, 08290, Cerdanyola del Vallés, Barcelona, Spain b Instituto de Ciencia de Materiales de Barcelona (CSIC), Campus UAB, 08193, Bellaterra, Spain c Departamento de Química, Universidad Autónoma de Barcelona, 08193, Bellaterra, Barcelona, Spain A R T I C L E I N F O A B S T R A C T Keywords: Two copper(II) complexes, both involving the anionic 3,5-bis(trifluoromethyl)benzoate ligand (TFMBz), have Dynamic porosity been prepared and their structure elucidated by single crystal-X ray diffraction. [Cu(TFMBz)2(ISNA)2] (1) is a Porous molecular crystals mononuclear complex, in which the sphere of coordination of Cu(II) is completed by two neutral isonicotinamide Adsorption (ISNA) auxiliary ligands. Hydrogen bonding formed between auxiliary ISNA ligands determines the formation of Selectivity a 2D supramolecular network. [Cu2(TFMBz)4(DMSO)2] (2) is a binuclear complex, in which the four carboxylate ligands define the typical paddle wheel structure often found in copper complexes, containing also the ancillary dimethyl sulfoxide (DMSO) ligand in axial positions. The elucidated crystallographic data provide the static view of the crystal structures, which reveals only non-interconnected voids for both materials. Even that, compound 1 shows an appreciable adsorption of CO2 at 273 K (ca. 1 CO2 molecule/Cu atom at 100 kPa), concomitant with a reduced adsorption of Ar or N2 under similar conditions, which implies a considerable degree of selectivity for CO2. Moreover, 1 does not adsorb N2 or Ar at 77 K and 100 kPa. This behavior suggests that the stablished intermolecular hydrogen bonds rest flexibility and dynamism to 1 at low temperature. Contrarily, increasing the temperature transient porosity is originated, which allows guest molecules to diffuse through the cavities. Compound 2 shows adsorption of N2 and Ar at 77 K, indicating that vibrations in the network and rotation of some CF3 groups, necessary to favor adsorbate diffusion, are still feasible at this low temperature. 1. Introduction materials based on discrete (instead of polymeric) units, which has recently gained considerable attention to shape highly functional porous Microporous materials have been the object of considerable research material. The development of this class of materials is still in an early due to the key roles that their pores play in many industrial processes. stage, with even no consensus in their terminology, interchangeably Although at industrial level the use of materials based on microporous receiving the name of porous molecular crystals (PMCs) [6–8], nano­ carbonaceous materials and zeolites is still highly predominant, several porous molecular crystals (NMCs) [9], hydrogen bonded organic alternative emergent microporous compounds are reaching interest. frameworks (HOFs) [10], hydrogen bonded organic-inorganic frame­ Among them, metal-organic frameworks (MOFs) are being explored in works (HOIFs) [11], and so on. The different terms are used to empha­ deep, due to the large number of parameters that can be chosen to tune size the nature of the discrete molecular units (organic or the properties of their pores, thus given place to thousands of different organometallic, i.e., coordination complexes) and/or to indicate the type materials [1–3]. Covalent organic frameworks (COFs) [4,5] form of intermolecular forces that connect the molecular units in the solid another family of microporous materials based in pure organic pre­ state. Along this work, the newly synthetized compounds are named as cursors. Both kind of materials, MOFs and COFs, have in common a PMCs, which we considered a comprehensive denomination. The prin­ structure based on a polymeric network, usually bi or tridimensional, cipal advantage of PMCs is solution processability, so the material can be defined by strong covalent bonds, including coordination bonds in redissolved for guest release and/or recrystallized for further reutiliza­ MOFs. tion. There are some examples of described PMCs with noticeable sur­ An important, but less researched, area is the study of microporous face area [12], potential for noble gasses separation and enantioselective * Corresponding author. E-mail address: Joseantonio.Ayllon@uab.es (J.A. Ayllón). https://doi.org/10.1016/j.micromeso.2023.112688 Received 13 April 2023; Received in revised form 6 June 2023; Accepted 7 June 2023 Available online 7 June 2023 1387-1811/© 2023 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). O. Vallcorba et al. Microporous and Mesoporous Materials 360 (2023) 112688 partitioning [13], or even responsiveness to external stimuli [14]. The 2.2. Synthesis porosity in PMCs may have different origins. In some cases, extrinsic porosity is derived from the inefficient packing of intrinsically 2.2.1. Synthesis of [Cu(TFMBz)2(ISNA)2] (1) non-porous molecules [15], while other PMCs are straightforwardly HTFMBz (1.032 g, 4.00 mmol) and Cu(CH3COO)2⋅H2O (0.399 g, built from intrinsically porous molecules, usually named as cages. When 6.02 mmol) were dissolved in methanol (100 mL) and, then, the proper the packing of these cages is inefficient, the resulting material has two amount of ISNA (0.500 g, 4.09 mmol) was added. The resulting blue sources of porosity, intrinsic and extrinsic [16,17]. Moreover, some solution was concentrated at room temperature in a laboratory hood molecular materials without intrinsic voids but presenting adsorption in until a blue-lavender solid crystallized. The solid was filtered, washed contact with specific guest molecules have been described [18]. These twice with a small volume of methanol (2x5 mL) and dried on air. Yield materials are prone to structural modifications, forced by the presence of was about 75 wt%. E.A. calcd. (wt%) for C30H18O6N4F12Cu (822.0 g the adsorbate, which generate pores and allows adsorption. These sub­ mol− 1): C, 43.83; H, 2.21; N, 6.82. Found: C, 43.93; H, 2.21; N, 6.56. stances are known as switching adsorbent molecular materials [19]. To attain PMCs built from intrinsically non-porous molecules, the 2.2.2. Synthesis of [Cu2(TFMBz)4(DMSO)2] (2) shape of the molecules is a key aspect, as some molecular geometries are HTFMBz (1.032 g, 4.00 mmol) and Cu(CH3COO)2⋅H2O (0.399 g, more prone to inefficient packing during crystallization than others 6.02 mmol) were added to DMSO (50 mL) and the suspension was [20–23]. However, the packing behavior is hard to anticipate, as inter­ heated to boiling. During this process, the initial greenish solution, with molecular forces (hydrogen bonding, π-π interactions and all sort of some solid in suspension, gave place to a blue solution. On standing at weak interactions) are also crucial in dictating the growth of the su­ room temperature, a blue crystalline solid precipitated, which was pramolecular network by crystallization. filtered, washed twice with a small volume of methanol (2x5 mL) and In intrinsically porous PMCs, the molecular units can either pack to dried on air. The yield was about 87 wt%. E.A. calcd. (wt%) for give a permanent interconnected porous structure or reveal the exis­ C40H24O10F24S2Cu2 (1311.8 g mol− 1): C, 36.62; H, 1.84. Found: C, tence of isolated, guest inaccessible, voids [24]. It is worth mentioning 36.63; H, 1.80. that usually porosity analyses are performed from the static perspective given by crystal structure data, not accounting by molecular vibrations 2.3. Characterization methods and rotation movements. For this reason, experimental adsorption data can noticeably differ from that simulated from structural data. As a Samples chemical composition was assessed by elemental analysis result, in some cases certain guest molecules can experimentally diffuse (E.A., Thermo Carlo Erba Flash 2000). Powder X-ray diffraction patterns inside a material structurally defined as nonporous. Hence, to under­ (PXRD) were recorded at room temperature in a Siemens D5000 stand the physicochemical properties of these materials it becomes diffractometer, using the Cu Kα incident radiation. Synchrotron powder necessary to take into account the dynamic porosity, since, under proper X-ray diffraction (SPXRD) measurements were performed at the BL16- experimental conditions, molecular and lattice flexibility can produce NOTOS beamline of the ALBA synchrotron. Data were collected at 20 “transient” channels that allow the diffusion of adsorbates through the keV (0.62074 Å wavelength) in the transmission mode using the host solid crystal structure [24–26]. Dynamic motion of some atoms or Mythen-II detector (6 modules at 535 mm from the sample). The molecular constituents of the solid host framework must be considered powdered sample was placed inside of an ad hoc capillary reaction cell to understand in which way the pores become interconnected to allow made of a fused silica capillary with 0.7 and 0.85 mm inner and outer adsorbate diffusion. This phenomenon has been termed by Barbour and diameters, respectively. The capillary was connected to both a gas- others as ‘‘porosity without pores’’ [27,28]. The necessary flexibility in handling and a vacuum line. An Oxford Cryostream 700 instrument the molecules forming the solid framework is activated thermically or was used to control the temperature of the sample. Differential scanning with other external stimuli, such as pression. In PMCs, the supramo­ calorimetry (DSC) experiments were performed under N2 atmosphere in lecular network flexibility is favored by the fact that the discrete a dynamic Mettler Toledo DSC 821e equipment. A thermal heat program building units are only connected by weak intermolecular forces. was settled consisting in cooling the sample from 25 to - 140 ◦ C at a Clearly, the understanding of the diffusion of guests within porous cooling rate of 5 ◦ C min− 1. After 1 min at this temperature, the reverse molecular solids needs to be increased to design new materials for heating at the same rate was performed. Molar enthalpies of transition specific applications. Hence, the main aim of this work is to contribute to (ΔH) were calculated from peak areas using original software provided the interpretation of these materials based on the analysis of two specific by Mettler Toledo. The gas adsorption behavior was determined at products involving bis(trifluoromethyl)benzoate (TFMBz)2 and Cu(II). either 77 or 273 K with N2, Ar and CO2 probe gasses using an ASAP 2000 Molecular units including the TFMBz ligand are prone to show ineffi­ Micromeritics Inc apparatus. Samples were first degassed at 373 K cient packing due to the disposition of the two voluminous tri­ during 20 h. fluoromethyl groups. This behaviour has previously been observed and reported for three different Co(II) 1D coordination polymers containing 2.4. Crystal structures this benzoate ligand [29]. In this work, two molecular Cu(II) complexes including TFMBz ligand and two different ancillary ligands (dime­ The crystal structures of 1 and 2 were solved using synchrotron thylsulfoxide (DMSO) and isonicotinamide (ISNA)) are synthetized and single-crystal XRD (SSCXRD) data collected in the XALOC beamline of described in regard of the crystal structure. N2, Ar and CO2 adsorption the ALBA Synchrotron [30]. Measurements were performed at 100 K experiments at different temperatures were performed to determine with a 0.72929 Å wavelength using the Dectris Pilatus 6 M detector which potential guest and at what temperature can diffuse into the voids placed at 120 mm from the sample. Three sets of 360⁰ φ-scans (in steps of of the solid material. 0.5⁰ and 0.15 s⋅step− 1) at three κ angles (0, 45 and 90⁰) were attempted and the successful ones were merged afterwards to increase data 2. Experimental section completeness and redundancy. Data were indexed, integrated and scaled using XDS software [31]. The crystal structures were solved by intrinsic 2.1. Materials phasing and refined with SHELXL [32] by full-matrix least-squares methods on F2 (version 2014/7) using Olex2 as the graphical interface Copper(II) acetate hydrate (Cu(CH3COO)2⋅H2O), 3,5-bis(trifluoro­ [33]. Hydrogen atoms bonded to carbon atoms were placed in calcu­ methyl)benzoic acid (HTFMBz), ISNA, DMSO and methanol were pur­ lated positions with isotropic displacement parameters fixed at 1.2 times chased from commercial sources and used as received. the Ueq of the corresponding carbon atoms. Data for compound 2 was also taken at room temperature using Mo Kα radiation (λ = 0.71073 Å) 2 O. Vallcorba et al. Microporous and Mesoporous Materials 360 (2023) 112688 in a SMART-APEX diffractometer with a CCD detector. An empirical TFMBz ligand are slightly tilted (7.15(10)⁰) respect the CuO4 plane. The absorption correction was applied (SADABS) and was solved and refined distorted octahedral geometry around each metal ion is completed by using the same procedure described previously. The crystallographic two pyridinic nitrogen atoms from two equivalent ISNA ligands (Cu-N data and refinement details for both compounds are summarized in 1.9942(18) Å). The plane defined by the aromatic ring of ISNA is almost Tables S1–S5 of the SI. orthogonal to the CuO4 plane (dihedral angle 89.96(9)⁰) (Fig. 1a). The CCDC 2240642–2240644 files contain the supplementary crystallo­ amide group of each ISNA establishes intermolecular hydrogen bonds graphic data for this paper. These data can be obtained free of charge via that connect two close ISNA ligands belonging to two different mole­ www.ccdc.cam.ac.uk/data_request/cif. cules (Fig. 1b, Table 1). One of the H atoms of each amide group is hydrogen bonded to the O of the amide group of a neighbor ligand (N(2)- 3. Results and discussion H(2A)⋅⋅⋅O(3)). These hydrogen bonds create the typical supramolecular symmetric amide-amide synthon [34,35]. Besides, the other H in each 3.1. Synthesis and basic characterization amide group is connected to one O atom of a carboxylate group in a TFMBz belonging to a third molecule (N(2)-H(2B)⋯O(2)). This In this work, two molecular solids were prepared by exchange re­ connection is reinforced by another contact between this O and one of actions in solution, in which the acetate present in the precursor was the H atoms of the aromatic ring of ISNA (C(12)-H(12)⋅⋅⋅O(2)). There­ replaced by the TFMBz ligand with the concomitant addition of ancillary fore, each molecular subunit is connected by hydrogen bonds to another neutral ligands ISNA or DMSO and the subsequent crystallization of [Cu four, two through each ISNA ligand. The expansion of these contacts (TFMBz)2(ISNA)2] (1) and [Cu2(TFMBz)4(DMSO)2] (2), respectively. generates supramolecular layers parallel to de ab plane (Fig. 1c). In these The synthesis of both compounds is straightforward from the HTFMBz layers, the TBMBz ligand protrude (Fig. 1d) and when the layers stack, and Cu(II) acetate reagents. The synthesis of (1) was carried out at room they interdigitate. The strongest inter-layer interactions are stablished temperature in methanol, which is a good solvent for both main reagents between the H of the aromatic ring at para position and the O of the and the ISNA ancillary ligand. Compound (2) was synthetized in DMSO, amide group of a close by layer, C(12)-H(12)⋅⋅⋅O(2) (Fig. 1e and f). solvent that was incorporated to the structure as axial ligand. In the Fluoride atoms do not participate in any remarkable intermolecular later, a small amount of an amorphous unknown insoluble intermediate interaction. This 3D packing determines non-interconnected voids, was formed at room temperature, reason why the mixture was heated to which accounts for a 9.3% of the crystal cell volume when measured boiling to assure the complete dissolution of the precursors. The desired with a probe radius of 1.2 Å using Mercury software (only 6.2% if the product crystallized on cooling. The elucidated crystal structure by probe radius used is 1.6 Å) (Fig. 2). SSCXRD for these compounds is described in the next section. Phase purity of the different precipitated compounds was first confirmed by E. 3.2.2. Structure of [Cu2(TFMBz)4(DMSO)2] (2) A., in which obtained C,H,N values matched the given stoichiometry in both cases. Second, PXRD patterns recorded and simulated from single 3.2.2.1. Low temperature phase (100 K). Compound 2 also belongs to crystal data were compared displaying perfect coincidence the triclinic crystal system, with the space group P-1. The asymmetric (Figs. S1–S2). Finally, the retention of crystallinity upon gas sorption unit contains a Cu(II) atom, two independent TFMBz anionic ligands and and desorption was confirmed by PXRD. one DMSO neutral ligand linked to the copper atom trough the oxygen 3.2. Structures Table 1 3.2.1. Structure of [Cu(TFMBz)2(ISNA)2] (1) Intermolecular interactions for [Cu(TFMBz)2(ISNA)2] (1). Compound 1 belongs to the triclinic crystal system, with the space D-H⋅⋅⋅A D-H (Å) H⋅⋅⋅⋅A (Å) D⋅⋅⋅⋅A (Å) > D-H⋅⋅⋅A (◦ ) group P-1. The asymmetric unit contains a Cu(II) atom, located at an N(2)-H(2A)⋅⋅⋅O(3) 0.86 2.03 2.887(3) 175 inversion center, a TFMBz anion, and one ISNA ligand. In this mono­ N(2)-H(2B)⋯O(2) 0.86 2.04 2.885(3) 168 nuclear complex, two equivalent TFMBz ligands coordinate the metal C(5)-H(5)⋅⋅⋅O(3) 0.93 2.34 3.243(3) 164 center in a strongly asymmetrical quelating mode (Cu-O distances of C(12)-H(12)⋅⋅⋅O(2) 0.93 2.25 3.157(3) 166 1.9764(14) and 2.5493(16) Å). The plane of the aromatic ring of the Fig. 1. Structure of 1, [Cu(TFMBz)2(ISNA)2]: (a) molecule detail; (b) detail of the hydrogen bonds stablished by each ISNA ligand; (c) 2D supramolecu­ lar network generated by the expansion of the former hydrogen bonds (view along c axis); and (d) lateral view of the 2D supramolecular network. Hydrogen atoms, except those participating in intermolecular hydrogen bonds, are omitted for clarity. Depicted hydrogen bonds are denoted as dotted green lines; (e) detail of the intermolecular contacts between adja­ cent supramolecular 2D networks; and (f) view of the packing highlighting the former contacts. 3 O. Vallcorba et al. Microporous and Mesoporous Materials 360 (2023) 112688 Fig. 2. Voids, determined with a probe radius of 1.2 Å (Mercury 2022.3), in the crystal structure of 1 [Cu(TFMBz)2(ISNA)2]: (a) view along b axis; and (b) view along c axis. atom. This binuclear complex adopts the typical paddle wheel structure: two equivalent Cu(II) are held together by four sin-sin μ2,η1,η1 carbox­ ylate bridges from four TFMBz ligands, while two DMSO occupy the axial positions. All the Cu-O(carboxylate) bond lengths are in a narrow interval (1.964(2)-1.975(2) Å), while the Cu-O(DMSO) bond length is Fig. 3. Crystal structure of 2, [Cu2(TFMBz)4(DMSO)2]: (a) molecule detail; (b) 2.101(4) Å (Fig. 3a). DMSO ligands of neighbor molecules are paired by detail of the stablished hydrogen bonds between adjacent molecules. Their a double hydrogen bond between H atoms of the methyl group and the O expansion defines chains parallel to the a axis; (c) view of the packing along the atom (O5⋯H20A-C20, dO⋅⋅⋅H 2.54 Å; dO⋅⋅⋅C = 3.386(7) Å), noticeably a axis; π-π interactions are depicted as dotted blue lines. For clarity, hydrogen atoms, except those participating in the hydrogen bonds, are omitted for clarity stronger than those observed in pure solid DMSO at the same temper­ and only the position most populated of each disordered fluoride atoms ature [36]. The expansion of these intermolecular hydrogen bonds de­ is depicted. fines supramolecular chains parallel to the a axis (Fig. 3b). These chains are interconnected by π - π intermolecular contacts in the [010] direction The former description is based on the determination of the structure (centroid-centroid distances 3.556(4) Å), combined with weaker π - π from crystallographic data measured at 100 K. Additional information interactions in the [001] direction (centroid-centroid distances 3.888(4) can be acquired by measuring the variable temperature SPXRD patterns Å) (Fig. 3c). of the sample. Measurements performed in the temperature interval 300 As two of the four independent -CF3 groups are disordered (two to 100 K indicated a sharp variation on the pattern at around 245 K alternative positions for each fluoride atom), guest accessible space (Fig. 5). This behavior is confirmed by DSC characterization, in which visualization is not trivial. When one of the fluoride positions is fixed to two close exothermic events are displayed at 256 and 248 K (Fig. 6). The the most populated position, only non-interconnected voids are gener­ correspondent enthalpy values are ca. − 0.7 and − 1.3 kJ/mol, respec­ ated. Contrarily, when the less populated positions are arbitrarily fixed tively. These thermal events correspond to a first-order phase transitions and the probe radius is unrealistically reduced to 1.1 Å, channels are [37]. To gain further information, structure elucidation has also been generated (Fig. 4). In fact, the fluoride disorder suggests that the energy done from data obtained at room temperature. Despite the changes barrier to the rotation of half of the -CF3 groups is still feasible at this low observed in the SPXRD patterns at ca. 245 K, the structure determined at temperature, and probably also at 77 K, the temperature used for the N2 room temperature is similar to the one described at 100 K. The spatial adsorption-desorption isotherm measurements. group was preserved, though, as expected due to thermal expansion, the cell volume increased in ca. 4.5%, while the angles varied in opposite 3.2.2.2. Phase change evidence and room temperature crystal structure. 4 O. Vallcorba et al. Microporous and Mesoporous Materials 360 (2023) 112688 Fig. 4. Crystal packing of [Cu2(TFMBz)4(DMSO)2]: view along c axis. Isolated voids appear with a probe radius of 1.2 Å, while the narrow channels turn up only by reducing the probe radius to 1.1 Å (Mercury 2022.3). Fig. 5. SPXRD patterns at variable temperature measured under vacuum for [Cu2(TFMBz)4(DMSO)2] (2). directions, e.g., alfa increased while beta and gamma decreased. The key accommodate a CO2 molecule, the analysis of the crystal structure does difference is that, at room temperature, all -CF3 groups, and also the not show interconnection between voids to allow the diffusion of any DMSO axial ligands, present disorder: two alternative positions for each gas. A plausible explanation to this disparity is that at 77 K the hydrogen fluoride atom in -CF3 groups and two alternative positions for all the bond supramolecular network was rigid enough to prevent the diffusion atoms of DMSO, except the oxygen atom linked to the metal cation. of any molecule between voids. On the contrary, at 273 K the thermal Hence, considering variable temperature SPXRD and DSC data, 2 shows energy was enough to foster a significant level of flexibility, including a phase change at ca. 245 K that involves variations in the order or vibration and rotation motions, even when that implies transient disorder of these parts of the molecule, half of the CF3 and DMSO, while hydrogen bonding weakening or rupture. This flexibility creates tran­ the other half of -CF3 still present disorder at low temperature. sient communication between initially isolated pores, allowing the diffusion of potential guest molecules [38,39]. The isosteric heats of adsorption (Qst) for compound 1 were measured as a function of CO2 3.3. Adsorption properties uptake (Fig. 8 and S3). The Qst values varied between − 28.8 kJ mol− 1 at low coverages and − 22.8 kJ mol− 1 when the amount of adsorbed CO2 Compound 1 does not show any appreciable adsorption of N2 at 77 K, increases. This range of values should be considered as moderate when denoting that the voids determined in the elucidated crystal structure compared to the Qst values reported for standard conventional porous are not accessible to guests. Conversely, 1 shows an appreciable materials applied for CO2 adsorption, such as microporous carbons, adsorption of CO2 at 273 K (Fig. 7), which corresponds approximately to zeolites or MOFs [40]. The obtained Qst value for 1 denotes phys­ one CO2 molecule per copper atom, or which is the same, one CO2 isorption, thus facilitating the reverse desorption process and the reuse molecule per void. Although the volume of each void is large enough to 5 O. Vallcorba et al. Microporous and Mesoporous Materials 360 (2023) 112688 of the materials [41]. Nevertheless, the adsorption of N2 or Ar at 273 K was noticeable lower than that of CO2, denoting a significant degree of selectivity for 1 (Fig. 7). It is also remarkable that, despite the high solvent accessible volume of 133 Å3 per unit cell according to Platon report [42], the as-synthesized 1 do not contain occluded solvent mol­ ecules (methanol or water), which also indicates a certain degree of selectivity towards potential guest. Compound 2 shows adsorption of N2 and Ar at 77 K (Fig. 9). The SSCXRD data analysis indicates the presence of channels, but only when using a probe radius of 1.1 Å, instead of the isolated voids observed in 1. The need of such a small probe radius to define the channels implies that the necks are not wide enough to allow the diffusion of N2 [43,44]. However, experimental measurements of the adsorption isotherm of N2 in 2, repeated several times and with different batch samples, evidenced adsorption of this gas at 77 K. Again, this implies that dynamic effects (vibrational and rotation motions) must account for the momentary widening necessary to adsorbate diffusion between voids [45]. The amount of N2 adsorbed corresponds roughly to one N2 molecule per void observed in Fig. 9a, or what is the same, one N2 per adsorbate molecular unit. The measured amount of adsorbed N2 at 100 kPa was 1.11 N2/Cu dimer, which take into account the additional contribution to adsorption of the external surface of the adsorbate particles. It is remarkable that in Fig. 6. DSC curves recorded for [Cu2(TFMBz)4(DMSO)2] (2) (black and red lines for cooling and heating, respectively, 5 K min− 1) indicating phase tran­ sitions. Peaks maximum/minimum are marked by arrows. Fig. 7. Adsorption/desorption isotherms of CO2 (red), Ar (blue) and N2 (black) at 273 K for [Cu(TFMbz)2(ISNA)2] (1). Fig. 8. Calculated Qst values at different surface coverages of CO2 adsorbed on [Cu(TFMbz)2(ISNA)2] (1). Fig. 9. Adsorption/desorption isotherms of N2 (a) and Ar(b) at 77 K for [Cu2(TFMBz)4(DMSO)2] (2). 6 O. Vallcorba et al. Microporous and Mesoporous Materials 360 (2023) 112688 this case the volume of the void per unit cell (46.1 Å3 per unit cell ac­ adsorption/desorption of different gases (CO2 on 1, and N2 or Ar on 2) cording Platon report) [38] was very close to the volume of a N2 denotes that in both PMCs different guests can diffuse through the molecule estimated in the liquid state (51.7 Å3). For Ar, a slightly larger structure, in spite that only isolated voids can be observed in the analysis amount than for N2 was adsorbed, in the order of 1.3 Ar/Cu dimer at of porosity performed from structural data. Experimental observations 100 kPa. The previously described disorder of half of the -CF3 groups in can only be rationalized considering dynamic porosity. For 2, the hys­ the structure at 100 K suggests that to cross the energy barrier for teresis loops and stepped shape of the N2 and Ar isotherms suggest that rotation of these -CF3 groups was still feasible at this low temperature, some small structural changes occurs concomitant with the adsorption and probably also at 77 K, the temperature of the N2 and Ar process, observation confirmed by SPXRD characterization. adsorption-desorption isotherms measurement. Rotation of some -CF3 groups, accompanied of some vibration, etc., must contribute to create CRediT authorship contribution statement the transient channels that allow N2 diffusion inside the solid material. The displayed N2 and Ar isotherms for 2 (Fig. 9) only reflects equilib­ Oriol Vallcorba: Methodology, Investigation, Formal analysis, Data rium positions. However, it is remarkable that the measurements take a curation. Julio Fraile-Sainz: Methodology, Investigation, Formal long time, longer than usual in a microporous compound. The fact that analysis. Concepción Domingo: Funding acquisition, Methodology, the time required to attain each equilibrium point is remarkably large, Project administration, Resources, Supervision, Investigation, Writing – suggests that the diffusion of the guest is somehow sluggish. Neverthe­ review & editing. José A. Ayllón: Writing – review & editing, Writing – less, the attained adsorption when related to the volume of the void is original draft, Supervision, Investigation, Formal analysis, close to the limit value. The amount of adsorbed CO2 at 273 was very Conceptualization. low, only 0.13 CO2/Cu dimer, and related to adsorption on the external surface of the material. The N2 and Ar adsorption isotherms recorded for 2 displayed some Declaration of competing interest unique characteristics (Fig. 9), such as hysteresis and stepped shape. Both effects have been associated with structural changes occurring The authors declare that they have no known competing financial during adsorption [46,47]. This behavior can also be observed interests or personal relationships that could have appeared to influence comparing the SPXRD spectra obtained under vacuum (on cooling in the work reported in this paper. Fig. 5) and atmospheric conditions (Fig. S4), displaying clear differences in the two diffraction peaks around 5.7◦ that appear doubled at the Data availability lower temperatures diffractograms. This difference is attributed to the presence of N2 in the experiment in which the sample holder was not Data will be made available on request. submitted to vacuum during cooling. The observed changes are a com­ bination of the effect of cooling and N2 adsorption, especially in the Acknowledgements lowest range of temperatures. In fact, the adsorption of N2 during the measurements also takes place during data acquisition of single crystal This work was supported by the Spanish Ministry of Science and X-ray diffraction at 100 K of, in which a flux of cool nitrogen is used to Innovation MICINN through the Severo Ochoa Program for Centers of control de temperature, being the monocrystal of 2 in direct contact with Excellence (CEX2019-000917-S) and the Spanish National Plan of it. The reported crystal structure was solved with the aid of the solvent Research with project PID2020-115631 GB-I00. The measurements for masking method (note the “[+solvent]” notation in the checkcif file). the elucidation of the crystal structures and SPXRD patterns were per­ Particularly, the electron density accounted with the solvent masking formed at the XALOC and NOTOS beamlines of the ALBA Synchrotron, methodology corresponds to six electrons by unit cell. As each N2 respectively, with the collaboration of ALBA staff. molecule has 14 electrons, these six electrons equal to 0.43 molecules, i. e., roughly half of the voids are filled with N2. Appendix A. Supplementary data This adsorption of N2 accounts for the difference between the simulated diffractogram from the single crystal data and the ones Supplementary data to this article can be found online at https://doi. measured under vacuum (Fig. S5, compare the black and red lines). The org/10.1016/j.micromeso.2023.112688. diffractogram of the sample cooled without applying vacuum (Fig. S5, blue line) displayed also some differences with the simulated one due to the slowness of the process of N2 adsorption, as discussed previously. References Nevertheless, when the sample is maintained under these conditions at [1] R. Freund, O. Zaremba, G. Arnauts, R. Ameloot, G. Skorupskii, M. Dincă, this temperature, the N2 adsorption continues and a diffractogram A. Bavykina, J. Gascon, A. Ejsmont, J. Goscianska, M. Kalmutzki, U. Lächelt, matching the simulated one is obtained (Fig. S5, green line). It is worth E. Ploetz, C.S. Diercks, S. Wuttke, The current status of MOF and COF applications, noting that, although the most intense peaks in the SSCXRD patterns do Angew. Chem. Int. Ed. 60 (2021) 23975–24001. [2] P. Silva, S.M.F. Vilela, J.P.C. Tomé, F.A. Almeida Paz, Multifunctional not experiment significative shifts, the described alterations of the lines metal–organic frameworks: from academia to industrial applications, Chem. Soc. around 5.7◦ clearly indicate some minor structural changes arising Rev. 44 (2015) 6774–6803. during the adsorption process. [3] Z. Ji, H. Wang, S. Canossa, S. Wuttke, O.M. Yaghi, Pore chemistry of metal–organic frameworks, Adv. Funct. Mater. 30 (2020), 2000238. [4] H. Lyu, H. Li, N. Hanikel, K. Wang, O.M. Yaghi, Covalent organic frameworks for 4. Conclusions carbon dioxide capture from air, J. Am. Chem. Soc. 144 (2022) 12989–12995. [5] M.S. Lohse, T. Bein, Covalent organic frameworks: structures, synthesis, and applications, Adv. Funct. Mater. 28 (2018), 1705553. Two new Cu(II)/TFMBz PMC complexes were synthetized and their [6] M.M. Deegan, M.R. Dworzak, A.J. Gosselin, K.J. Korman, E.D. Bloch, Gas storage in structure elucidated crystallographically: the mononuclear [Cu porous molecular materials, Chem. Eur J. 27 (2021) 4531–4547. (TFMBz)2(ISNA)2] (1) and the binuclear [Cu2(TFMBz)4(DMSO)2] (2). In [7] I. Bassanetti, A. Comotti, P. Sozzani, S. Bracco, G. Calestani, F. Mezzadri, L. Marchiò, Porous molecular crystals by macrocyclic coordination these two compounds, the type of established intermolecular forces and supramolecules, J. Am. Chem. Soc. 136 (2014) 14883–14895. energy determine the solid-state dynamics, and, particularly, the mini­ [8] H. Yamagishi, Functions and fundamentals of porous molecular crystals sustained mum temperature necessary to allow a dynamic motion of parts of the by labile bonds, Chem. Commun. 58 (2022) 11887–11897. structure (rotation, vibration, breaking and reforming hydrogen bonds, [9] N.B. McKeown, Nanoporous molecular crystals, J. Mater. Chem. 20 (2010) 10588–10597. etc), which is necessary to allow the diffusion of adsorbates by the [10] P. Li, M.R. Ryder, J.F. Stoddart, Hydrogen-bonded organic frameworks: a rising generation of an apparent microporous compound. The reversible class of porous molecular materials, Acc. Mater. Res. 1 (2020) 77–87. 7 O. Vallcorba et al. Microporous and Mesoporous Materials 360 (2023) 112688 [11] D. Shao, P. Peng, M. You, L.-F. Shen, S.-Y. She, Y.-Q. Zhang, Z. Tian, Hydrogen- coordination polymers built from 3,5-Bis(trifluoromethyl)benzoate and 4,4′ - bonded framework of a cobalt(II) complex showing superior stability and field- bipyridine, Cryst. Growth Des. 22 (2022) 4463–4471. induced slow magnetic relaxation, Inorg. Chem. 61 (2022) 3754–3762. [30] J. Juanhuix, F. Gil-Ortiz, G. Cuní, C. Colldelram, J. Nicolás, J. Lidón, E. Boter, [12] G. Zhang, O. Presly, F. White, I.M. Oppel, M. Mastalerz, A permanent mesoporous C. Ruget, S. Ferrer, J. Benach, Developments in optics and performance at BL13- organic cage with an exceptionally high surface area, Angew. Chem. Int. Ed. 53 XALOC, the macromolecular crystallography beamline at the Alba Synchrotron, (2014) 1516–1520. J. Synchrotron Radiat. 21 (2014) 679–689. [13] L. Chen, P.S. Reiss, S.Y. Chong, D. Holden, K.E. Jelfs, T. Hasell, M.A. Little, [31] W. Kabsch, XDS, Acta Crystallogr D Biol Crystallogr 66 (2010) 125–132. A. Kewley, M.E. Briggs, A. Stephenson, K.M. Thomas, J.A. Armstrong, J. Bell, [32] G.M. Sheldrick, Crystal structure refinement with SHELXL, Acta Crystallogr C J. Busto, R. Noel, J. Liu, D.M. Strachan, P.K. Thallapally, A.I. Cooper, Separation of Struct Chem 71 (2015) 3–8. rare gases and chiral molecules by selective binding in porous organic cages, Nat. [33] O.V. Dolomanov, L.J. Bourhis, R.J. Gildea, J.A.K. Howard, H. Puschmann, OLEX2 : Mater. 13 (2014) 954–960. a complete structure solution, refinement and analysis program, J. Appl. [14] H. Sasaki, H. Imoto, T. Kitao, T. Uemura, T. Yumura, K. Naka, Fluorinated porous Crystallogr. 42 (2009) 339–341. molecular crystals: vapor-triggered on–off switching of luminescence and porosity, [34] C.B. Aakeröy, A.M. Beatty, J. Desper, M. O′ Shea, J. Valdés-Martínez, Directed Chem. Commun. 55 (2019) 6487–6490. assembly of dinuclear and mononuclear copper(ii)-carboxylates into infinite 1-D [15] N. Dwarkanath, S. Palchowdhury, S. Balasubramanian, Unraveling the sorption motifs using isonicotinamide as a high-yielding supramolecular reagent, Dalton mechanism of CO 2 in a molecular crystal without intrinsic porosity, J. Phys. Chem. Trans. (2003) 3956–3962. B 123 (2019) 7471–7481. [35] F. Sánchez-Férez, L. Bayés, M. Font-Bardia, J. Pons, Solvent dependent formation [16] E. Sánchez-González, M.Y. Tsang, J. Troyano, G.A. Craig, S. Furukawa, Assembling of Cu(II) complexes based on isonicotinamide ligand, Inorg. Chim. Acta. 494 metal–organic cages as porous materials, Chem. Soc. Rev. 51 (2022) 4876–4889. (2019) 112–122. [17] M.J. Bojdys, M.E. Briggs, J.T.A. Jones, D.J. Adams, S.Y. Chong, M. Schmidtmann, [36] H. Reuter, Structural parameters of dimethyl sulfoxide, DMSO, at 100 K, based on a A.I. Cooper, Supramolecular engineering of intrinsic and extrinsic porosity in redetermination by use of high-quality single-crystal X-ray data, Acta Crystallogr E covalent organic cages, J. Am. Chem. Soc. 133 (2011) 16566–16571. Cryst Commun 73 (2017) 1405–1408. [18] S. Noro, T. Ohba, K. Fukuhara, Y. Takahashi, T. Akutagawa, T. Nakamura, Diverse [37] L. Yuan, S. Clevers, A. Burel, P. Negrier, M. del Barrio, B. Ben Hassine, D. Mondieig, structures and adsorption properties of quasi-Werner-type copper(ii) complexes V. Dupray, J.Ll Tamarit, G. Coquerel, New intermediate polymorph of 1-Fluoro- with flexible and polar axial bonds, Dalton Trans. 40 (2011) 2268–2274. adamantane and its second-order-like transition toward the low temperature [19] A.M. Kałuża, S. Mukherjee, S.-Q. Wang, D.J. O’Hearn, M.J. Zaworotko, [Cu(4- phase, Cryst. Growth Des. 17 (2017) 3395–3401. phenylpyridine)4 (trifluoromethanesulfonate) 2 ], a Werner complex that exhibits [38] A. Aguilar-Granda, S. Pérez-Estrada, E. Sánchez-González, J.R. Álvarez, high selectivity for o -xylene, Chem. Commun. 56 (2020) 1940–1943. J. Rodríguez-Hernández, M. Rodríguez, A.E. Roa, S. Hernández-Ortega, I.A. Ibarra, [20] C. Cui, P.R. Shipman, R.A. Lalancette, F. Jäkle, Tris(2-pyridylborate) (tpyb) metal B. Rodríguez-Molina, Transient porosity in densely packed crystalline carbazole–(p complexes: synthesis, characterization, and formation of extrinsically porous -Diethynylphenylene)–Carbazole rotors: CO 2 and acetone sorption properties, materials with large cylindrical channels, Inorg. Chem. 52 (2013) 9440–9448. J. Am. Chem. Soc. 139 (2017) 7549–7557. [21] B. Kohl, F. Rominger, M. Mastalerz, Crystal structures of a molecule designed not to [39] M. Inukai, M. Tamura, S. Horike, M. Higuchi, S. Kitagawa, K. Nakamura, Storage of pack tightly, Chem. Eur J. 21 (2015) 17308–17313. CO 2 into porous coordination polymer controlled by molecular rotor dynamics, [22] E. Coropceanu, A. Rija, V. Lozan, I. Bulhac, G. Duca, V.Ch Kravtsov, P. Bourosh, Angew. Chem. Int. Ed. 57 (2018) 8687–8690. Discrete binuclear cobalt(III) bis-dioximates with wheel-and-axle topology as [40] M. Oschatz, M. Antonietti, A search for selectivity to enable CO 2 capture with building blocks to afford porous supramolecular metal–organic frameworks, Cryst. porous adsorbents, Energy Environ. Sci. 11 (2018) 57–70. Growth Des. 16 (2016) 814–820. [41] M. Khraisheh, S. Mukherjee, A. Kumar, F. Al Momani, G. Walker, M.J. Zaworotko, [23] A. Bacchi, P. Pelagatti, Fabricating organometallic wheel-and-axle compounds for An overview on trace CO2 removal by advanced physisorbent materials, the creation of dynamically porous crystalline materials, CrystEngComm 18 (2016) J. Environ. Manag. 255 (2020), 109874. 6114–6123. [42] A.L. Spek, Single-crystal structure validation with the program PLATON, J. Appl. [24] D. Holden, S.Y. Chong, L. Chen, K.E. Jelfs, T. Hasell, A.I. Cooper, Understanding Crystallogr. 36 (2003) 7–13. static, dynamic and cooperative porosity in molecular materials, Chem. Sci. 7 [43] D. Ongari, P.G. Boyd, S. Barthel, M. Witman, M. Haranczyk, B. Smit, Accurate (2016) 4875–4879. characterization of the pore volume in microporous crystalline materials, Langmuir [25] J.L. Atwood, L.J. Barbour, A. Jerga, B.L. Schottel, Guest transport in a nonporous 33 (2017) 14529–14538. organic solid via dynamic van der Waals cooperativity, Science 298 (2002) [44] M. Wu, M. Li, L. Yuan, F. Pan, “Useless channels” in a molecular crystal formed via 1000–1002. F⋅⋅⋅F and F⋅⋅⋅π halogen bonds, Cryst. Growth Des. 22 (2022) 971–975. [26] S. Noro, Y. Meng, K. Suzuki, M. Sugiura, Y. Hijikata, J. Pirillo, X. Zheng, [45] T. Kurihara, M. Inukai, M. Mizuno, Slow CO 2 diffusion governed by steric K. Takahashi, T. Nakamura, A temporarily pore-openable porous coordination hindrance of rotatory ligands in small pores of a metal–organic framework, J. Phys. polymer for guest adsorption/desorption, Inorg. Chem. 60 (2021) 4531–4538. Chem. Lett. 13 (2022) 7023–7028. [27] L.J. Barbour, Crystal porosity and the burden of proof, Chem. Commun. (2006) [46] F.-X. Coudert, J.D. Evans, Nanoscale metamaterials: meta-MOFs and framework 1163–1168. materials with anomalous behavior, Coord. Chem. Rev. 388 (2019) 48–62. [28] L. Dobrzańska, G.O. Lloyd, H.G. Raubenheimer, L.J. Barbour, Permeability of a [47] Q. Yang, P. Lama, S. Sen, M. Lusi, K. Chen, W. Gao, M. Shivanna, T. Pham, seemingly nonporous crystal formed by a discrete metallocyclic complex, J. Am. N. Hosono, S. Kusaka, J.J. Perry, S. Ma, B. Space, L.J. Barbour, S. Kitagawa, M. Chem. Soc. 128 (2006) 698–699. J. Zaworotko, Reversible switching between highly porous and nonporous phases [29] F. Papi, A. Rosado, O. Vallcorba, A.E. Lanza, M. Gemmi, N. Portolés-Gil, A. of an interpenetrated diamondoid coordination network that exhibits gate-opening M. López-Periago, C. Domingo, J.A. Ayllón, Supramolecular isomerism in cobalt(II) at methane storage pressures, Angew. Chem. Int. Ed. 57 (2018) 5684–5689. 8