Copper(II) halide salts of 5-bromo-2-aminopyridine and 3,5-dibromo-2-aminopyridine: syntheses, structures and magnetic behavior

Abstract Two 5-bromo-2-aminopyridinium quantum Heisenberg antiferromagnetic (QHAF) complexes of copper(II) have been synthesized and studied structurally and magnetically. The structures of (5-BAPH)2[CuCl4] (1) and (5-BAPH)(3,5-diBAPH)[CuBr4] (2) are reported [5-BAPH = 2-amino-5-bromopyridinium; 3,5-diBAPH = 2-amino-3,5-dibromopyridinium]. Single-crystal X-ray diffraction measurements show that 1 crystallizes in the triclinic space group Pī and 2 crystallizes in the monoclinic space group P21/c. Both products crystallize with a four-coordinate Cu(II) ion and are stabilized with extensive hydrogen bonding between the halide ions and pyridinium and amino substituents on each cation. The tetrahalidocuprate ions in 1 and 2 exist in distorted geometries between square planar and tetrahedral with mean trans angles of 139.48° and 131.85°, respectively. Halogen bonding is also observed. Magnetic susceptibility data were collected on 1 and 2. Magnetic susceptibility data of 1 exhibit moderate antiferromagnetic interactions which are best fit using an alternating chain model [CC = 0.4262(3) cm3 K mol−1, J/kB = −22.75(2) K, α = 0.134(3)]. Magnetic susceptibility data of 2 exhibit weak antiferromagnetic interactions, best fit using a uniform chain model [CC = 0.410(6), J/kB = −6.1(4) K].


Introduction
Magnetism has been of special interest for decades, including discoveries in spinorientation [1,2], magnetic exchange [3][4][5], and spin-lattice relaxation [6,7].Interest in these fields demonstrates how vast and exciting magnetochemistry can be, especially when concerning coordination chemistry, magnetic structure, and potential magnetic properties.Transition metal complexes have been of particular interest in molecular magnetism due to their unique lattice structural properties and magnetic interactions, leading to magnetic exchange properties such as large anisotropy [8,9] and multiferroics [10,11].Renewed interest in antiferromagnetic copper(II) complexes can be attributed to the theory of high-temperature superconductivity in layered copper oxide complexes [12].Copper(II) has a S ¼ 1/2 magnetic moment and a d 9 electron configuration.Complexes of copper(II) ions are well-known to range from four-to sixcoordinate.Four-coordinate copper(II) complexes are typically observed in distorted tetrahedral geometries and can crystallize with up to 3-dimensional (3D) magnetic interactions.
With the inherent flexibility of the Cu(II) coordination sphere, a variety of structures can be formed by changing minor properties of the ligands and counterions.Theoretical calculations for magnetic exchange exist for two-dimensional quantum Heisenberg antiferromagnets (2D QHAF), for which there have been many published complexes [13][14][15][16][17][18].Following the introduction of superconductivity in copper(II) compounds, an increased number of theoretical data-fitting models for QHAFs were published and an increase in Cu(II) compound syntheses was observed.These moderate to weakly antiferromagnetic compounds, synthesized to test theoretical magnetic data fitting models, can show magnetic superexchange between copper(II) ions via coordinated bridging ligands, or the magnetic exchange may be propagated through two-halide superexchange pathways between coordinated halide ions [19].Several parameters between the tetrahalidocuprate ions have been identified that directly impact the strength of magnetic exchange (J) propagated through the two-halide pathway.Changes in the distance between the non-bonding halide ions, the Cu-XÁ Á ÁX angles, and the Cu-XÁ Á ÁX-Cu torsion angles [20] impact orbital overlap between the halide ions, affecting the magnetic exchange (Scheme 1).
Moderate magnetic exchange can be achieved using copper(II) salts such as tetrahalidocuprate ions.Extensive research has been dedicated to the study of complexes containing these ions [21][22][23], producing 2D crystal lattices with honeycomb [24], ladder [25], and layered [26] magnetic structures.Changes in the degree of delocalization of spin density from the Cu(II) ion also impacts the strength of the magnetic exchange in systems utilizing the two-halide pathway [27].The degree of delocalization of spin density is determined by the structure of the CuX 4 À2 ion, which could form tetrahedral or square planar geometries.However, the Jahn-Teller distortion prevents these Cu(II) complexes from obtaining idealized tetrahedral geometry and the degree of distortion can be described via the mean trans angle [28].
structures, and magnetic properties of (

Experimental
Copper(II) chloride dihydrate and copper(II) bromide were purchased from Mallinckrodt Pharmaceuticals.

Magnetic measurements
Magnetic data for 1 and 2 were collected on a Quantum Design MPMS SQUID Magnetometer.Powdered samples of 1 (26.4 mg) and 2 (53.1 mg) were placed in gelatin capsules which were mounted in plastic straws.Magnetization was measured in applied fields ranging from 0 to 50 kOe; several points were collected while the field was returning to 0 kOe to check for hysteresis effects; none were observed.M(H) was linear to at least 2 kOe for 1 and 2. Temperature-dependent magnetization data were collected from 1.8 K to 310 K in an applied field of 1 kOe.All data were corrected to account for the capsule and straw (measured independently), along with the diamagnetic contributions from constituent atoms, estimated via Pascal's constants [46] and the temperature-independent paramagnetism of the Cu(II) ion.Data were fit using the H ¼ ÀJ P S i ÁS j Hamiltonian.Powder X-ray diffraction data were collected to confirm the purity and phase of the material on a Bruker AXS-D8 Focus X-ray Powder Diffractometer (1

X-ray structure analysis
X-ray diffraction data for 1 and 2 were collected on a Siemens-Bruker P4/CCD diffractometer with monochromated Mo Ka radiation at 162 C and 168 C, respectively.Data collection, reduction and absorption corrections were made using Bruker APEX2, SAINT v.8.34A, and SADABS v.2016 [47].The structures were solved by direct methods using SHELXS-2014 [48] and refined via least squares analysis with SHELXL-2018/3 [49].Anisotropic thermal parameters were used to refine the positions of non-hydrogen atoms.Hydrogen atoms bonded to carbon atoms were placed in geometrically calculated positions and refined using a riding model with fixed isotropic thermal parameters.Hydrogen atoms bonded to nitrogen atoms were located in the difference map and allowed to refine with fixed isotropic thermal parameters.Full crystal and refinement details are given in Table 1 and relevant bond lengths and angles are given in Table 2. Hydrogen bonding lengths and angles are given in Table 3.The crystal structure data have been deposited with the CCDC: 1, 2240748; 2, 2240747.
The original synthesis of 2 is suspected to be the result of bromination of the 5-BAPH aromatic ring, synthesizing 3,5-diBAPH in solution.This is a result that was unexpected, but not unprecedented.Similar bromination has been observed in aminopyridine cations under comparable conditions [50,51].Complexes were analyzed by IR (Figures SI-3 (1) and SI-4 (2)), variable temperature magnetic susceptibility, and single crystal X-ray diffraction.The bromine substituents are also only slightly distorted from the plane of the ring at 0.037 Å (Br1) and 0.027 Å (Br2).The bond lengths and angles within the N11 and N21 rings are normal within experimental error compared to the previously reported pyridinium cation [37].In spite of conjugation, there is a 16.4 deviation of the plane of the amino group (H22A-N22-H22B) from the plane of the N21 aromatic  ring.This deviation is likely the result of the N22-H22B … Cl3 hydrogen bond (see Table 3).The organic cations are held in the lattice by hydrogen bonding between the pyridinium hydrogen atoms of the 5-BAPH and the chloride ions coordinated to the Cu(II): N11-H11Á Á ÁX (X ¼ Cl1, Cl3) and N21-H21Á Á ÁX (X ¼ Cl2, Cl3) as well as hydrogen bonds involving the amino groups.Figure 2 shows the hydrogen bonding in the unit cell.
Compound 2 crystallizes in the monoclinic space group P2 1 /c with four molecular units per unit cell.The asymmetric unit is shown in Figure 3.
The mean trans-Br-Cu-Br angle of 131.85 illustrates the distorted tetrahedral geometry of the Cu(II) ion [52].Bond lengths and angles of the 5-BAPH and 3,5-diBAPH aromatic rings agree with previously reported structures for 5-BAPH [43] and 3,5-diBAPH [44].The 5-BAPH and 3,5-diBAPH rings are both virtually planar with mean deviations of 0.0285 Å and 0.0479 Å, respectively.The bromine substituents on each organic molecule are slightly distorted from the aromatic ring plane with the largest deviation being 0.060 Å (Br15).There are significant deviations of the amino substituents from the aromatic ring plane of 7.96 and 13.17 , observed in the 5-BAPH and 3,5-diBAPH molecules, respectively.The larger deviation observed in the 3,5-diBAPH amino substituent may result from the N12-H12A … Br4 hydrogen bond (see Table 3), similar to that observed in 1.The counterions are stabilized within the lattice through significant hydrogen bonding as observed in 1 (Figure 4).
Many of the same supramolecular interactions present in 1 are seen in the crystal structure of 2. Bifurcated hydrogen bonding contributes significantly to the stabilization of the lattice and is observed on the 5-BAPH pyridinium hydrogen atom in 1 and the 3,5-diBAPH pyridinium hydrogen atom in 2.
Halogen bonding is present for further crystal structure stabilization between halogenated cations and tetrahalidocuprate anions.These bonds can exist between C- YÁ Á ÁX-M (Y, X ¼ Cl, Br and M ¼ transition metal cation) or between two metal-halide cations.The YÁ Á ÁX distances are typically less than the sum of the van der Waals radii [53]. 1 exhibits halogen bonding between the chloride ions (Cl3 and Cl4) and each bromine atom on the N11 and N21 cations (Figure 5).There are no halogen bonds between two bromine substituents in 1.

Magnetic properties
Field-dependent data at 1.8 K for 1 display downward curvature as shown in Figure SI-5.M(H) is linear from 0 to 2000 Oe and hysteresis was not observed.Compound 1 reaches a maximum of 65.4 emu mol À1 at 50 kOe, which is far below the expected saturation magnetization for Cu(II) compounds ($6000 emu-mol À1 ) [46].This indicates that the bulk sample exhibits a singlet ground state and the data at 1.8 K represent a trace paramagnetic impurity.
Magnetic susceptibility as a function of temperature was measured from 1.8 K to 310 K in a 1 kOe field.The high-temperature 1/v mol (T) were fit to the Curie-Weiss model (Figure SI-6), resulting in a Curie constant of 0.421(1) cm 3 mol À1 and a Weiss constant, h ¼ À19.5(4) K, indicative of modest antiferromagnetic interactions.Figure 7 displays the sharp, rounded maximum observed in the v mol (T) data, characteristic of a magnetic singlet ground state (Figure SI-7).
The maximum of v mol ¼ 1.51 Â 10 À3 cm 3 mol À1 occurs at T ¼ 14.1 K.A singlet ground state is observed, indicated by a steep downward curve following the maximum.The data were initially fit using an antiferromagnetic strong-rung spin ladder model [55] with magnetic interactions of J/k B (rung) À22.58(6)K and J/k B (rail) ¼ À1.7(4) K, a Curie constant of 0.431(3) cm 3 mol À1 , and a 1.40(3)% impurity (vide infra).With further inspection of the crystal structure, the v(T) data (Figure 7) and vT(T) data (Figure SI-8) were fit to an antiferromagnetic alternating chain model yielding a Curie constant of 0.426(1) cm 3 mol À1 , J/k B ¼ À22.8(2) K, a J 0 /J ratio (a) of 0.134(4) [J 0 /k B ¼ À3.04(4) K], and a 1.36(6)% impurity.Field-dependent data at 1.8 K for 2 display slight upward curvature, as shown in Figure SI-9, indicative of a low-dimensional magnetic lattice.M is linear from 0 to 20,000 Oe and hysteresis was not observed.Compound 2 reaches a maximum of 1330 mu mol À1 at 50 kOe, remaining well below the expected saturation magnetization for Cu(II) compounds ($6000 emu-mol À1 ).The magnetic susceptibility as a function of temperature data for 2 were measured and the high-temperature 1/v mol data were fit to the Curie-Weiss model (Figure SI -10), which resulted in a Curie constant of 0.4219(5) cm 3 mol À1 and a Weiss constant of À6.6(2) K, suggesting slightly weaker  antiferromagnetic interactions than 1.The v mol data were fit to the antiferromagnetic uniform chain model (Figure 8) [55].
The maximum of v mol ¼ 2.80 Â 10 À2 cm 3 mol À1 occurs at T ¼ 4.66 K.The data were fit to an antiferromagnetic uniform chain model [55] with a Curie-Weiss correction for inter-chain interactions.The vT(T) data were similarly fit (Figure .This resulted in a Curie constant of 0.410(6) cm 3 mol À1 , J/k B ¼ À6.1(4) K and a paramagnetic impurity of 6.1(2)%.The fitted Weiss constant (0.0(4) K) indicated that magnetic interactions between chains were negligible.Data were initially fit to an antiferromagnetic alternating chain model with magnetic interactions of J/k B ¼ À8.53(3) K and a ¼ 0.959 (4).Due to the symmetry of potential magnetic interactions indicated by the crystal structure, the absence of a singlet ground state, and the resulting a implying nearly equivalent magnetic exchange values, the alternating chain model was exchanged for the uniform chain model with better fitting results.
Although 1 and 2 exist in different space groups and contain different copper-halide ions, they both were predicted to propagate magnetic exchange through a honeycomb-like magnetic lattice.Parameters for potential magnetic exchange in 1 and 2, using the two-halide superexchange pathway, identified halide-halide contacts between 3.0 and 5.0 Å and suggested that both complexes would exhibit three close X … X contacts (Figures 9 and 10).
Through analyzing magnetic data fitting, a magnetic lattice can be proposed.Because there is no available QHAF model for a honeycomb lattice, an equivalent lattice model was adopted.Low-dimensional QHAFs exist with two to four close contacts within a magnetic lattice creating chain, ladder, honeycomb, and grid patterns [56].Both the spin ladder and honeycomb magnetic lattices contain three close contacts per copper-halide ion (Scheme 4).
If the two magnetic models are equivalent, the honeycomb model can be described in terms of rails (vertical interactions) and rungs (horizontal, alternating interactions).Halide-halide close contact lengths, angles, and torsion angles are reported for both lattices using the short contacts labeled in Figures 5 and 6 (Table 5).For every rung contact in a spin ladder model, there must be two rail contacts.The rails are required to have the same magnetic interaction strength (identical by symmetry) within the honeycomb lattice if they are to be considered analogous to a ladder model.For 1, there were three distinct interaction distances within the proposed honeycomb model, implying three different values for magnetic exchange.Thus, the proposed honeycomb magnetic lattice was not analogous to the spin ladder model.The process of assigning rungs and rails to each compound was straightforward for 2 as the Cu-Br1Á Á ÁBr2 interactions were distinct from the Cu-Br4Á Á ÁBr1 interactions.
The initial data fitting of the strong-rung spin ladder model for 1 was indicative of a magnetic lattice with only two J parameters, which appeared incompatible with the crystal structure due to the presence of three unique close contact distances.A singlet  ground state is observed for both the spin ladder [57] and alternating chain models [58], therefore the data were fit to an antiferromagnetic alternating chain with equally well-fit results.The Cu-Cl3Á Á ÁCl4 dimer close contact interaction (A) is the most likely negligible interaction in the magnetic lattice due to the much larger distance, 4.509 Å compared to the Cu-Cl1Á Á ÁCl1 (B) or Cu-Cl2Á Á ÁCl2 (C) close contacts, 3.584 Å and 3.806 Å, respectively.The contacts A, B, and C are shown in Figure 11.
Compound 2 contains similar halide-halide contacts to 1, creating chains with contact lengths of 3.829 Å (A), separated by a distance of 4.637 Å (B) with long dimer interactions (Figure 10).Following incompatible fitting to strong rung and strong rail spin ladder models, further examination of the structure suggested an alternating chain due to the presence of two unique close contact interactions, A and B interactions, and the data were fit to an antiferromagnetic alternating chain with interactions of J/k B ¼ À8.531(3) K and a ¼ 0.959(4).However, the ratio (a) indicated nearly equivalent magnetic interaction strengths and the alternating close contacts were symmetry equivalent.The data showed a better fit using the antiferromagnetic uniform chain model.The results indicate that 2 exists in an antiferromagnetic uniform chain magnetic lattice and the Cu-Br4Á Á ÁBr1 (B) interaction is negligible (Figure 10).

Conclusion
The two-halide superexchange pathway provides for magnetic exchange and analysis of the crystal structure allows for magnetic lattice prediction.The magnetic data provide for modification of that prediction based upon the combination of magnetic behavior and the symmetry exhibited in the crystal structure.Although both 1 and 2 Scheme 4. Magnetic exchange between metal ions in a ladder (black and blue) and a honeycomb (black and red) lattice.were predicted to exhibit a magnetic honeycomb lattice, analysis of the magnetic data indicated that neither compound conformed to that magnetic lattice, one being an alternating chain and the other a uniform chain.The change from the predicted magnetic lattices indicates clearly that additional compounds, and magnetic data, are needed to provide a more quantitative approach to understanding the two-halide superexchange pathway.Further, the lack of a theoretical model for the honeycomb lattice calls on the theory community to develop a QHAF honeycomb model as more compounds are synthesized, allowing for greater in-depth magnetic analysis.

Figure 1 .
Figure 1.The asymmetric unit of 1 showing a 50% probability thermal ellipsoids.Hydrogen atoms are shown as spheres of arbitrary size and only those whose positions were refined are labeled.

Figure 2 .
Figure 2. A plot showing the unit cell with hydrogen bonds (dashed lines) via chloride ions in 1.

Figure 3 .
Figure 3.The asymmetric unit of 2 showing a 50% probability thermal ellipsoids.Hydrogen atoms are shown as spheres of arbitrary size and only those whose positions were refined are labeled.

Figure 4 .
Figure 4.A plot showing the unit cell viewed parallel to the a-axis with hydrogen bonds (dashed lines) and the linking via hydrogen bonds and bromide ions in 2.

Figure 7 .
Figure 7.The powder magnetic susceptibility vs. temperature for 1.The same figure is provided on a logarithmic scale in Figure SI-7.

Figure 8 .
Figure 8.The powder magnetic susceptibility vs temperature for 2. The solid line represents the best fit to the uniform S ¼ 1 = 2 antiferromagnetic chain model.

Figure 9 .
Figure 9.The proposed honeycomb magnetic lattice for 1. Dashed lines represent short halide-halide contacts.

Figure 10 .
Figure 10.The proposed honeycomb magnetic lattice for 2. Dashed lines represent short halidehalide contacts.

Figure 11 .
Figure 11.The potential magnetic honeycomb lattice of 1. Magnetically active close contacts are shown as dashed lines and close contacts not involved in superexchange are shown as dotted lines.
, giving a colorless solution.CuBr 2 (0.119 g, 0.53 mmol) was dissolved in 5 mL of H 2 O to yield a turquoise solution.The CuBr 2 solution was added to the combined 5-BAP/3,5-diBAP solutions, resulting in an immediate color change to dark green.The reaction mixture was allowed to evaporate slowly at room temperature.After four months, the filtrate produced large, dark purple, rectangular, prismatic crystals which were isolated by filtration and dried mechanically.Yield, 0.313 g (89.5%).

Table 1 .
Crystal data and structure refinement parameters for 1 and 2.