Supramolecular isomerism in cobalt(II) coordination polymers built from 3,5-bis(trifluoromethyl)benzoate and 4,4´-bipyridine. Francesco Papi,a Albert Rosado,b Oriol Vallcorba,c Arianna E. Lanza,a Mauro Gemmi,a,* Núria Portolés-Gil,b Ana M. López-Periago,b Concepción Domingo,b and José A. Ayllón.d,* a Istituto Italiano di Tecnologia, Center for Materials Interfaces, Electron Crystallography, Viale Rinaldo Piaggio 34, 56025 Pontedera, Italy b Instituto de Ciencia de Materiales de Barcelona (ICMAB-CSIC), Campus UAB s/n, Bellaterra 08193, Spain c ALBA Synchrotron Light Source, Cerdanyola del Vallés, Barcelona, Spain d Departamento de Química, Universidad Autónoma de Barcelona, 08193-Bellaterra, Barcelona, Spain Supporting Information 1 Characterization techniques p. 3 Table S1. Crystallographic data for compounds 1, 2, 3, 5 determined by single p. 8 crystal diffraction Table S2. Crystallographic data for compound 4 determined by powder XRD p. 9 Figure S1. ATR-FTIR spectra of compounds 1–5 p. 10 Figure S2. Binuclear Co2(TFMBz)4 secondary building units and ladder chains p. 11 observed in the crystal structures of compounds 2 (a) and 4 (b Figure S3. Details of chain packing in compound 2 p. 12 Figure S4. Details of chain packing in compound 3 p. 13 Figure S5. Details of chain packing in compound 4 p. 14 Figure S6. Void space analysis of the crystal structures of compounds 2, 3, 4 p. 15 Figure S7. SEM micrograph of compound 2 p. 16 Figure S8. Thermal variation of 1/χ for compounds 1–5 at an applied field of 1000 p. 17 Oe Figure S9-S13. Powder X-ray diffractograms for compound 1–5 p. 18 Figure S14. Fit of the rigid-body Rietveld refinement for compound 4 against p. 20 PXRD data 2 Characterization Techniques Single crystal X-ray and electron diffraction Single-crystal X-ray diffraction (SCXRD) experiments for the structural elucidation of compound 1 were performed in the XALOC beamline at the ALBA synchrotron (Spain).1 Data were collected at 100 K with a 0.72931 Å wavelength using the Dectris Pilatus 6M detector placed at 120 mm from the sample. The ϕ-scans were performed from 0 to 360° in steps of 0.5° with a collection time of 0.15° step-1. The scan was repeated at three different κ angles (0, 45 and 90°), and merged afterwards to increase the completeness and redundancy when possible. Data were indexed, integrated and scaled using the XDS software.2 The crystal structure was solved by intrinsic phasing and refined with SHELXL (version 2014/7)3 using Olex2 as graphical interface.4 Crystallization trials for compounds 2–5 yielded micrometric crystals that were not suitable for single crystal X-ray diffraction but conversely were good candidates for 3D electron diffraction (3D ED). Crystals of these compounds were gently crushed and directly loaded on carbon-coated Cu TEM grid without any solvent. 3D ED experiments were performed at room temperature on a Zeiss Libra 120 TEM operating at 120 kV and equipped with a LaB6 thermionic source. 3D ED patterns were collected from micrometric single crystals on an ASI Timepix single electron detector using a 1° precessed electron beam (NanoMEGAS DigiSTAR P1000 device) of ∼150 nm obtained through a 5 µm C2 condenser aperture. The data were analyzed and processed by using Pets5 and indicated a C-centered orthorhombic cell for compound 2, a triclinic cell for compounds 3 and 4 and a C-centered monoclinic cell for compound 5 (Table S1). Ab initio methods successfully provided the structural solution for compound 2 in space group C m c a (charge-flipping method, Superflip)6 and compound 5 in space group C 2/c (direct methods, SIR2019),7 while compound 3 structure was solved in space group P -1 with simulated annealing (SIR2019) using as a starting guess of the molecular model the unpublished structure of zinc- derived compound 3. As regards the structure of compound 4, the sample inevitably suffered 3 from electron beam sensitivity and displayed highest resolution of 2.0 Å, even when cooled at 100 K. Nevertheless, data collected at room temperature allowed us to unambiguously determine the lattice parameters for its triclinic unit cell. Moreover, simulated annealing procedures (SIR2019), carried out as for compound 3, delivered a consistent solution for the SBU in space group P -1. Because of the limited data quality, any refinement trial was not successful, and the partial molecular model was then manually improved and validated against powder X- ray diffraction (see next section). The structural models were kinematically refined against 3D ED data by Shelxl8,9 using electron atomic scattering factors. More accurate unit cell parameters provided by powder X-ray diffraction (see next section) were used in the refinement in order to improve the structures quality. Suitable geometrical ties were added according to the Grade Web Server10 and implemented using the DSR software.11 The hydrogen atoms were generated in geometrically idealized positions and all atoms were refined isotropically in the refinement procedures. Due to the low completeness and high number of parameters refined versus 3D ED data, merging of datasets collected from five single crystals of compound 3 was performed in Jana200612 and the resulting reflection file was used for the final refinement. Upon inspection of the difference electron (electrostatic) density maps, no solvent molecule entrapped in the voids of compounds 2 and 3 was possibly localized. The analysis of the crystal structure has been carried out thanks to the Mercury software.13 Data on single crystal X-ray and electron diffraction of compounds 1, 2, 3, 5 is summarized in Table S1. Powder X-ray diffraction Powder X-ray diffraction patterns (PXRD) for phase identification were recorded at room temperature for compound 1–5 in a Siemens D5000 diffractometer, using Cu Kα incident 4 radiation and a Bragg-Brentano geometry (Figure S9–S13). All samples were previously activated under vacuum (degassed at 333 K during 20 h) and grounded by hand in an agate pestle. Powder XRD data on compounds 2–5 were also acquired in Debye-Scherrer geometry using a STOE Stadi P diffractometer equipped with Cu-Kα1 radiation (λ = 1.5406 Å), a Ge (111) Johansson monochromator from STOE & Cie and a MYTHEN2 1K detector from Dectris. The same samples used for 3D ED analyses were loaded in a borosilicate glass capillary (0.2 mm external diameter) and data were acquired in the 2θ range 2.5–50° (maximum resolution ca. 1.5 Å) with an interval of 0.015° between consecutive points. The unit cell and structural parameters were refined with Jana2006.12 In all cases the quality of PXRD data was limited by X-ray fluorescence of the Co atoms and by the fact that all compounds diffracted to low resolution, hence ruling out the possibility to perform complete Rietveld refinements. For all compounds, lattice parameters and peak profiles were refined, while the structure models were kept as semirigid bodies. Not only did PXRD data allow us to validate the structure of compounds 2, 3, 5 solved and refined from 3D ED data but also to elucidate the structure of compound 4 (Figure S14) For this latter, 3D ED data allowed only to determine the unit cell and the position of “Co2(TFMBz)4” dimers. Based on our previous knowledge on the intra-chain geometry, we could manually complete the model and then refine it against PXRD data. The weak and broad 001 peak at 5.65 ° was excluded from refinement, because a simple pseudo-Voigt function with anisotropic peak broadening was not sufficient to ensure a proper fit. Due to the limited data quality available we chose to not overparameterize the refinement and not make use of sophisticated empirical corrections (e.g. spherical harmonics). Pyridyl rings and TFMBz fragments were described as rigid bodies and the intramolecular geometry (e.g. bpy and carboxylic groups torsion angles) could not be refined to convergence, due to the limited data quality and resolution (ca 1.82 Å). The oscillations of atomic parameters can also be indicative of linker disorder, as also observed in 2. Importantly, despite the above limitations the structure refined by powder data is almost completely superimposable 5 with the partial 3D ED model and is in agreement with the experimental PXRD pattern, therefore can be considered a plausible model for compound 4 (see Table S2). Chemical composition, porosity and magnetic properties Samples chemical composition was assessed by elemental analysis (E.A., Thermo Carlo Erba Flash 2000). Surface area and porosity properties were determined by N2 adsorption/desorption at 77 K using an ASAP 2000 Micromeritics Inc. equipment. The micropore specific surface area was determined by the t-plot method. For both characterizations, samples were first degassed at 333 K during 20 h to eliminate adsorbed solvent molecules. Crystal morphology was observed by scanning electron microscopy (SEM) in a QUANTA FEI 200 FEG-ESEM microscope. Magnetic measurements were performed from 2 to 300 K in a Quantum Design MPMS-5S SQUID susceptometer. The molar susceptibility was corrected for the sample holder and for the diamagnetic contribution of all atoms by means of Pascal’s tables.14 References (1) Juanhuix, J.; Gil-Ortiz, F.; Cuní, G.; Colldelram, C.; Nicolás, J.; Lidón, J.; Boter, E.; Ruget, C.; Ferrer, S.; Benach, J.J. Synchrotron. Rad. 2014, 21, 679-689. (2) Kabsch, W. Acta Cryst. 2010, D66, 125-132. (3) Sheldrick, G. M. Acta Cryst. 2015, C71, 3-8, https://doi.org/ 10.1107/S2053229614024218 (4) Dolomanov, O. V.; Bourhis, L. J.; Gildea, R. J.; Howard, J. K.; Puschmann, H. J. Appl. Cryst. 2009, 42, 339-341. (5) Palatinus, L.; Brázda, P.; Jelínek, M.; Hrdá, J.; Steciuk, G.; Klementová, M. Acta Cryst. 2019, B75, 512– 522. (6) Palatinus, l.; Chapuis, G. J. Appl. Cryst., 2007, 40, 786-790. 6 (7) Burla, M.C.; Caliandro, R.; Carrozzini, B.; Cascarano, G.L.; Cuocci, C.; Giacovazzo, C.; Mallamo, M.; Mazzone, A.; Polidori. G. J. Appl. Cryst. 2015, 48, 306–309. (8) Sheldrick, G.M. Acta Cryst., 2015, C71, 3–8. (9) Hübschle, C.B.; Sheldrick, G.M.; Dittrich. B. J. Appl. Cryst., 2011, 44, 1281–1284. (10) O.S. Smart, T.O. Womack, A. Sharff, C. Flensburg, P. Keller, W. Paciorek, C. Vonrhein, and G. Bricogne. grade v.1.2.9 (2011) Cambridge: Global Phasing Ltd. http://www.globalphasing.com (11) Kratzert, D.; I. Krossing, I. J. Appl. Cryst., 2018, 51, 928–934. (12) V. Petříček, M. Dusek, and L. Palatinus. Z. Kristallogr., 2014, 229, 345–352. (13) C.F. Macrae, I. Sovago, S.J. Cottrell, P.T.A. Galek, P. McCabe, E. Pidcock, M. Platings, G.P. Shields, J.S. Stevens, M. Towler, and P.A. Wood. J. Appl. Cryst., 2020, 53, 226–235. (14) Bain, G. A., Berry, J. F.. J. Chem. Educ., 2008, 85, 532-536. 7 Table S1. Crystallographic data for compounds 1, 2, 3, 5 determined by single crystal diffraction. Compound [Co2(H2O)(TFMBz)4(py)4] 1 [Co2(TFMBz)4(bpy)2]n 2 [Co2(TFMBz)4(bpy)2]n 3 [Co(TFMBz)2(bpy)]n 5 Empirical formula C56H34F24N4O9Co2 C112H56Co4F48N8O16 C56H28Co2F24N4O8 C28H14CoF12N2O4 Formula weight 1480.73 2917.37 1458.69 729.34 Radiation X-rays Electrons Electrons Electrons T (K) 100(2) 303(2) 303(2) 303(2) Wavelength (Å) 0.72931 0.0335 0.0335 0.0335 System, space group Triclinic, P -1 (2) Orthorhombic, C m c a (64) Triclinic, P -1 (2) Monoclinic, C 2/c (15) a (Å) 13.506(2) 11.3742(4) 11.5123(7) 25.2067(6) b (Å) 14.5030(9) 34.8620(15) 15.0828(16) 11.3529(2) c (Å) 16.1460(14) 17.5715(5) 18.3730(7) 9.9530(2) α (°) 69.8010(5) 90 102.118(8) 90 β (°) 76.171(7) 90 90.946(7) 94.426(2) γ (°) 82.743(3) 90 98.648(9) 90 V (Å3) 2878.8(6) 6967.6(4) 3079.9(4) 2839.75(10) Z 2 2 2 4 Dcalc (g cm-3) 1.708 1.391 1.573 1.706 µ (mm-1) 0.761 0.011 0.012 0.013 F(000) 1480 937 468 468 Crystal size (mm3) 0.080x0.040x0.020 nanocrystal nanocrystal nanocrystal hkl ranges -18≤h≤18, -20≤k≤20, -24≤l≤24 -10≤h≤10, -28≤k≤27, -15≤l≤15 -10≤h≤10, -14≤k≤14, -16≤l≤17 -21≤h≤20, -10≤k≤10, -8≤l≤8 2θ range (°) 1.410 to 34.098 0.078 to 0.878 0.053 to 0.922 0.093 to 0.878 Reflections collected/unique/[Rint] 105281/18225/[Rint = 0.0414] 5040/1302/[Rint = 0.2849] 9631/4911/[Rint = 0.1540] 2260/763/[Rint = 0.1931] Completeness 83.2% 86.7% 86.0% 67.0% Refinement method Full matrix least-squares on |F|2 Full matrix least-squares on |F|2 Full matrix least-squares on |F|2 Full matrix least-squares on |F|2 Data/restrains/parameters 18225/2/865 1302/190/188 4911/380/377 763/116/95 Goodness of fit (GOF) on |F|2 1.175 1.093 1.103 1.104 Final R indices [I>2σ(I)] R1 = 0.0576, wR2 = 0.1450 R1 = 0.2697, wR2 = 0.6081 R1 = 0.3369, wR2 = 0.6942 R1 = 0.2999, wR2 = 0.6384 R indices (all data) R1 = 0.0578, wR2 = 0.1453 R1 = 0.3356, wR2 = 0.6665 R1 = 0.4078, wR2 = 0.7441 R1 = 0.3342, wR2 = 0.6805 Extinction coefficient 0.289(5) / / / Largest. Diff. peak and hole (e Å-3) 1.210 and -1.800 0.233 and -0.194 0.391 to -0.306 0.441 and -0.348 8 Table S2. Crystallographic data for compound 4 determined by powder XRD. Compound [Co(TFMBz)2(bpy)]n 4 Empirical formula C28H14CoF12N2O4 Formula weight 729.34 Radiation X-rays T (K) 298 Wavelength (Å) 1.5406 System, space group Triclinic, P -1 (2) a (Å) 10.2180(13) b (Å) 10.5716(12) c (Å) 16.098(3) α (°) 100.944(11) β (°) 93.532(11) γ (°) 113.522(5) V (Å3) 1547.1(4) Z 2 Dcalc (g cm-3) 1.5657 µ (mm-1) 5.554 F(000) 726 Crystal size (mm3) powder 2θ range (°) collected (step) 2.497–50.005 (0.015) 2θ range (°) used for refinement 7.51–50.005 Number of points collected 3168 Goodness of fit (GOF) 2.8530 RBragg 0.1965 R(F2) 0.1223 Rwp 0.0184 Rp 0.0133 Rexp 0.012 0.0065 9 Figure S1. ATR-FTIR spectra of compounds 1–5. 10 Figure S2. Binuclear Co2(TFMBz)4 secondary building units and ladder chains observed in the crystal structures of compounds 2 (a) and 4 (b). 2 contains disordered bpy and TFMBz ligands, not shown for clarity. Color code: C gray; O red; N blue, Co pink, and F green. H atoms are omitted . 11 Figure S3. Chain packing in compound 2. For clarity, only one position for each disordered TFMBz and bpy ligands is depicted. a) Code name of the chains. In all the A and A’ chains, the cobalt atoms lie in the same bc planes. For B and B’ chains, cobalt atoms lie in an intermediate position respect to A (or A’) chains. A and A’ differ in the orientation of “Co2(TFMBz)4” SBU. The same applies for B and B’. Color code: C gray; O red; N blue, Co pink, and F green. H atoms are omitted b) Intertwining of neighbor chains along c direction. c) Intertwining of neighbor chains along b direction. 12 Figure S4. Chain packing in compound 3. a) Code name of the chains. In all the A and A’ chains, the cobalt atoms lie in the same bc planes. For B and B’ chains, cobalt atoms lie in an intermediate position respect to A (or A’) chains. A and A’ differ in the orientation of “Co2(TFMBz)4” SBU. The same applies for B and B’. b) Intertwining of neighbor chains along c direction. c) Intertwining of neighbor chains along b direction. Color code: C gray; O red; N blue, Co pink, and F green. H atoms are omitted . 13 Figure S5. Chain packing in compound 4. a) Code name of the chains. In all the A1 chains, the cobalt atoms lie in the same plane perpendicular to the chain axis. In B1 chains, cobalt atoms lie in an intermediate position respect to A1. b) Detail of the packing of neighbor chains “A1-B1- A1”. c) Detail of the packing of neighbor chains “A1-A2-A3”. Color code: C gray; O red; N blue, Co pink, and F green. H atoms are omitted . 14 Figure S6. Void space analysis of the crystal structures of compounds 2 (a), 3 (b), 4 (c). Voids calculated with a probe radius of 1.2 Å. The void volume corresponds to 22.2 %, 15.1% and 6.5 % for compounds 2, 3 and 4, respectively. For compound 2, the voids are evaluated neglecting the structural disorder of bpy and TFMBz ligands (i.e. one of the two possible positions was chosen for each disordered bpy and TFMBz linker). 15 Figure. S7. SEM micrograph of compound 2 16 . Figure S8. Thermal variation of 1/χ (black dots are for the data used to fit to the Curie-Weiss equation) for compounds 1–5 at an applied field of 1000 Oe. 17 Figure S9. (Red line) X-ray diffractogram of [Co2(H2O)(TFMBz)4(py)4] (1) measured at room temperature, (Blue line) Calculated pattern from resolved crystal structure from monocrystal XRD measured at 100 K. Small displacement are due to different temperatures used to measure the powder and the single crystal samples. Figure S10. (Red line) X-ray diffractogram of [Co2(TFMBz)4(bpy)2]n (2) measured at room temperature. (Blue line) Calculated pattern from resolved crystal structure. 18 Figure S11. (Red line) X-ray diffractogram of [Co2(TFMBz)4(bpy)2]n (3) measured at room temperature. (Blue line) Calculated pattern from resolved crystal structure. Figure S12. (Red line): X-ray diffractogram of [Co2(TFMBz)4(bpy)2]n (4) measured at room temperature. (Blue line) Calculated pattern from resolved crystal structure. 19 Figure S13. (Red line) X-ray diffractogram of [Co(TFMBz)2(bpy)]n (5) measured at room temperature. (Blue line) Calculated pattern from resolved crystal structure. Figure S14. Fit of the rigid-body Rietveld refinement for compound 4 against PXRD data (Cu K1) collected in transmission mode (for details see section Materials and methods). The high background is due to X-ray fluorescence by Co atoms due to the wavelength used. Black: experimental pattern; red: calculated pattern; blue: difference; red markers: position of Bragg peaks; gray area: excluded region. 20