Engineered π⋯π interactions favour supramolecular dimers X@[FeL3]2 (X = Cl, Br, I): solid state and solution structure

Ditopic bis-pyrazolylpyridine ligands usually react with divalent metal ions (M2+) to produce dinuclear triple-stranded helicates [M2L3]4+ or, via π⋯π interactions, dimers of monoatomic complexes ([ML3]2)4+. The introduction of an additional benzene ring at each end of ligand L increases the number of aromatic contacts within the supramolecular aggregate by 40%, driving the self-recognition process in an irreversible manner. Consequently, the mixing of new bis-pyrazolylquinoline L2 with FeX2 salts leads to crystallization of the tripartite high-spin assemblies (X@[Fe(L2)3]2)3+ (X = Cl, Br or I). The aggregates exhibit exceptional stability, as confirmed by a combination of paramagnetic 1H NMR techniques, demonstrating their persistence in solution. Our investigations further reveal that the guests Br− and I− are retained inside the associate in solution but Cl− is immediately released, resulting in the formation of the empty supramolecular dimer ([Fe(L2)3]2)4+.

1,3-bis-(1-(quinolin-2-yl)propane-1,3-dione)-benzene (L'2).A solution of t BuOK (16mmol) in dry THF (16ml) was added dropwise and with stirring to a solution of 1,3-diacetophenone (0.65g, 4mmol) in dry THF (100ml) under N2.Upon further stirring for 15 min, the solution changed to light-yellow colour.Methyl quinoline-2carboxylate (1.5g, 8mmols) 1 was then added and the solution was brought to reflux overnight.During this process the mixture turned dark yellow.THF was then evaporated under low pressure, yielding a reddish-brown colour that was rinsed with water and then suspended in a biphasic solution of water (200ml) and diethyl ether (100ml).37% HCl was added until pH=2 was reached and the mixture was stirred for a few minutes, verifying that the pH remained at 2-3.After this time, the formation of an abundant white precipitate was observed.The latter was filtered under vacuum and 1.49 g of white solid (79.2%) were obtained.

Physical characterization
Elemental Analysis.C, H, N analyses were performed by using a Thermo EA Flash 2000 (Thermo Scientific) analyser at the Centres Científics i Tecnològics from the Universitat de Barcelona (CCiT-UB).
Infrared Spectroscopy.IR spectra were recorded as KBr pellet samples on a Nicolet 5700 FTIR spectrometer.
Magnetometry.Variable-temperature magnetic measurements were performed using a Quantum Design SQUID evercool MXL7 magnetometer at the "Unitat de Mesures Magnètiques" of the Universitat de Barcelona.The diamagnetic contributions to the susceptibility were corrected using Pascal's constant tables.Further corrections of the sample holders contribution, determined empirically, were also applied.Variable temperature dc data were collected with an applied field of 5000 Oe, in settle mode.

Figure S6
. Representation of the (X@[Fe(L2)3]2) 3+ supramolecular cations of compounds 1, 2 and 3 (X = Cl, 1; Br, 2; I, 3), down the axis containing X and the Fe centres.X is purple, Fe is yellow and the ligands for each complex are thick green and thin black sticks, respectively.H atoms are not shown.

Experimental SCXRD
Data for compounds 1 and 2 were collected at 100 K at Beamline 12.2.1 of the Advanced Light Source (Berkeley, USA), on a Bruker D8 diffractometer equipped with a PHOTON II detector and using silicon (111) monochromated synchrotron radiation (λ = 0.7288 Å).Data for compound 3 were acquired at 100 K on the BL13-XALOC beamline 2 of the ALBA synchrotron (λ = 0.72932 Å).Data reduction and absorption corrections for 1, and 2 were performed with respectively SAINT and SADABS. 3Data reduction for compounds 3 was done with autoproc package 4 and XDS. 5 All structures were solved by intrinsic phasing with SHELXT 6 and refined by full-matrix least-squares on F 2 with SHELXL. 7Free refinement of the position of hydrogens atoms of the pyrazole N-H group did not converge, probably due to the too faint corresponding electron density peak present in difference Fourier maps.These hydrogens were therefore refined with their thermal parameter 1.5 times that of their carrier N atom and a distance restraint at 1.0 angstrom, based on the structure of free pyrazole reported using both X-ray and neutron diffraction. 8The rest of hydrogens were placed geometrically and refined with a riding model.In the structures of 1, 2 and 3, lattice solvent molecules were too diffuse/disordered to be modelled satisfactorily so they have been treated as a diffuse contribution to the overall scattering without specific atom positions by PLATON/SQUEEZE. 9l details can be found in CCDC 2279862-2279863-2279864 (1-2-3), which contain the supplementary crystallographic data for this paper.These data can be obtained free of charge from The Cambridge Crystallographic Data Center via https://summary.ccdc.cam.ac.uk/structure-summary-form.Crystallographic and refinement parameters are summarized in Table S1, Table S2 provides Fe-N bond lengths while Tables S3 and S4, details of hydrogen bonds and stacking interactions, respectively.Table S3.Hydrogen bonds involved in the supramolecular X@[Fe(L2)3]2 assemblies in the structures of compounds 1, 2 and 3.

D-H•
. Plots of χT vs T for compounds 1 (Cl), 2 (Br) and 3 (I), in the 2 (Cl and Br) or 4 (I) to 300 K temperature range.The value of the plateau is above that expected for two isotropic Fe 2+ centers in the HS state (S = 2) with g = 2 and no spinorbit coupling contributions (6 cm 3 Kmol −1 ).The observed data may be subject to small errors ascribed to the attributed molar mass (dependent on the exact amount of solvation molecules or moisture).The drop at lower temperatures is expected for this ion, subject to zero field splitting effects and the consequent failure of the Curie law in this regime.To minimize the effect of the paramagnetic relaxation, the diffusion coefficients were calculated using the signals of the (Br@[Fe(L2)3]2) 3+ moiety in the range 8.5 -10.5 ppm and of free ligand in the range 7.4 -8.6 ppm, followed by averaging the S/S0 intensity ratio.The diffusion coefficient was determined as slope of a tangent line in coordinatesln(S/S0) vs (γGzδ) 2 (Δ-δ/3).The calculated diffusion coefficients were D = 9.66 cm 2 /s for free ligand L2 and D = 5.89 cm 2 /s for the species (Br@[Fe(L2)3]2) 3+ .

Figure S11.
Comparison of aromatic diamagnetic region of 1 H spectrum of compound 2, X = Br (bottom) with the one for the free ligand L2 (top).Conditions: CD3CN, 400 MHz, room temperature.The sample of compound 2 was obtained by preliminary washing the single crystals of 2 with tetrahydrofuran (a good solvent for free ligand that does not dissolve the complex 2) with further drying and dissolution in CD3CN.
The ratio of the concentrations of the compound 2 and free ligand L2 in solution was estimated based on the comparison of integrals of pyrazole protons of complex (δ = 6.29 ppm, I = 6.00) and the ligand (δ = 7.56 ppm, I = 1.01).As the ligand has twofold symmetry, absent in the complex, and each complex molecule has six ligands, the resulting 6-to-1 ratio of integrals corresponds to ratio of 2:1 of molar concentration of 2 to L2 in solution.In other words, only one out of each thirteen L2 molecules is free in solution, and the rest are present as the part of the complex.S5.Assignment of the paramagnetic NMR spectra at room temperature for compounds 1, 2 and 3.

H NMR experimental data
1 H and 1 H-1 H COSY NMR spectra were recorded from solutions in acetonitrile-d3 with a Bruker Avance 600 and Bruker Avance 300 FT-spectrometer (600.2MHz and 300.1 MHz 1 H frequency) using the residual signals of the solvent ( 1 H 1.94 ppm) as internal standards.To calibrate the temperature within the temperature range 190 -300 K, a Bruker standard temperature calibration sample (4% of MeOH in methanold 4 ) was used.Above 300 K, the temperature was calibrated using a known dependence of the chemical shifts of pure ethylene glycol.The DOSY spectrum was obtained using a standard bipolar pulse sequence (ledbpgp2s) and the following parameters: acquisition time 3 s, relaxation delay 1 s, diffusion time 50 ms, gradient pulse length 1.25 ms.Diffusion coefficients were determined as the slope of tangent lines in coordinates -ln(S/S0) vs (γGzδ) 2 (Δ-δ/3).

Computational Details
The energy gain upon formation of the dimers ([Fe(L1)3]2) 4+ and ([Fe(L2)3]2) 4+ was evaluated after fully optimizing their geometry and that of the corresponding monomers.These structural optimizations were carried out using density functional theory (DFT) based calculations employing the PBE 10 exchange-correlation functional and a def2-SVP 11 basis set using the Gaussian 16 code. 12We have opted for using the PBE functional, instead of a hybrid functional (e.g.B3LYP), as a compromise between computational cost and accuracy.It is worth mentioning that PBE has already been successfully used to model systems with FeII ions in a highspin state. 13Spin-unrestricted calculations were done when dealing with all the species because the high-spin of the Fe(II) ions were considered.In all the calculations, the DFT-D3(BJ) semi-empirical dispersion potential introduced by Grimme 14 was added to the conventional Kohn-Sham energy for a proper description of the van der Waals interactions.Calculations in acetonitrile were performed using the polarizable continuum model. 15
Figure S3.Molecular representation of the (Cl@[Fe(L2)3]2) 3+ assembly of 1. Unique heteroatoms are labelled.Grey balls are C, white balls are H (only H atoms from N−H gro ups are shown).

Figure S4 .
Figure S4.Molecular representation of the (Br@[Fe(L2)3]2) 3+ assembly of 2. Unique heteroatoms are labelled.Grey balls are C, white balls are H (only H atoms from N−H groups are shown).

Figure S5 .
Figure S5.Molecular representation of the (I@[Fe(L2)3]2) 3+ assembly of 3. Unique heteroatoms are labelled.Grey balls are C, white balls are H (only H atoms from N−H groups are shown).

Figure S12. 1 H
Figure S12. 1 H 2D EXSY spectrum of compound 2; X=Br (CD3CN, 400 MHz) at room temperature.The weak blue cross peaks are NOE-type dipolar cross-peaks between the signals of the protons of the free ligand.There are no exchange cross-peaks with the same sign as diagonal peaks, implying the absence of the exchange phenomena in the NMR timescale.

Table S1 .
Crystallographic and refinement parameters for the structures of compounds 1, 2 and 3.