Crystal structures of 5,12-dimethyl-1,4,8,11-tetraazacyclotetradecane cobalt(III) mono-phenylacetylide and bis-phenylacetylide

The crystal structures presented herein consist of two positively charged CoIII(DMC) acetylide complexes that take on a pseudo-octahedral symmetry and can be synthesized under weak-base conditions.


Structural commentary
Compound 1 crystallizes in P1 with two crystallographically independent moieties, Fig. 1. Each moiety consists of one complex [Co III (DMC)(C 2 Ph)(Cl)] + cation, a chloride counterion, and one acetonitrile and methanol solvate molecule, for a total composition of C 20 H 33 ClCoN 4 ÁC 2 H 3 NÁCH 4 OÁCl. The two unique moieties, labeled A and B, are related by a pseudoglide plane (see the Supramolecular features section for a more detailed discussion), and a common atom-naming scheme was used for the contents of the two unique halves of the structure. Both methanol molecules are disordered, with a common refined occupancy ratio of 0.643 (16):0.357 (16).
Compound 2 crystallizes in P2 1 , Fig. 2. Similar to 1, 2 also features two unique cations and anions in its asymmetric unit, but they are not related by any crystallographic pseudosymmetry. Each complex cation [Co III (DMC)(C 2 Ph) 2 ] + is paired with a triflate anion. The asymmetric unit is completed by a single methylene chloride solvate molecule, yielding a formula of 2(C 28 H 38 CoN 4 )Á2(CF 3 O 3 S)ÁCH 2 Cl 2 . One of the triflate anions as well as the methylene chloride molecule were refined as disordered, with occupancy rates of 0.503 (22) and 0.545 (12) for the major components.
The molecular geometries of the cations in 1 and 2 are similar (Tables 1 and 2). Both structures feature a central Displacement ellipsoid plot (50% probability setting) for compound 2, showing the atom-naming scheme and some of the hydrogen-bonding interactions (turquoise dashed lines). The OTf molecule comprising S2 is in position x À 1, y, z. The carbon-bound H atoms and H-atom labels are omitted for clarity.

Figure 1
Displacement ellipsoid plot (50% probability setting) for one of the two pseudo-symmetry-related halves of the asymmetric unit of compound 1, showing the atom-naming scheme and some of the hydrogen-bonding interactions (turquoise dashed lines). Shown is the 'B-moiety', the atomnaming scheme for the 'A-moiety' is equivalent. Labels for H atoms are omitted for clarity. cobalt(III) ion with a pseudo-octahedral geometry. The metal ion is coordinated in the equatorial plane by the four amine nitrogen atoms of a 5,12-dimethyl-1,4,8,11-tetraazacyclotetradecane (DMC) ligand. For compounds 1 and 2 respectively, the nearly linear C-Co-Cl [177.7 (2) and 178.0 (2) ] and C-Co-C [177.67 (9) and 179.67 (9) ] units are close to normal to the equatorial plane created by the coordinated amines of the macrocyclic ligand, confirming octahedral geometries. The C-Co-N and Cl-Co-N angles are all tightly clustered around 90 . The actual values range from 87.1 (1) to 92.9 (1) (Tables 1 and 2). The N-Co-N angles are more variable, caused by the difference in size of the ethylene and 1-methyl-propylene bridges of the DMC ligand. They range from 85.6 (3) to 94.4 (3) , with the smaller values being associated with the shorter ethylene N-CH 2 -CH 2 -N chelates, and the larger with the wider N-CH(CH 3 )-CH 2 -CH 2 -N connections (Tables 1 and 2).
Some of the Co-C C angles deviate from perfect linearity, likely due to steric forces resulting from packing effects. The values range from 171.3 (7) to 174.2 (2) , with the latter extreme value belonging to one of the Co-C C units of 2. All Co-C C angles are given in Tables 1 and 2. Each macrocycle exhibits a trans-III RRSS conformation, characterized by two neighboring N-H amine units pointing upwards, while their two trans N-H counterparts point in the opposing direction. Other conformations, such as trans-I, II, IV, or cis conformations, are much less prevalent for both the DMC and other cyclam ligands when coordinated to transition metals. (Bosnich et al., 1965;Hoffert et al., 2012;Cook et al., 2016).
The Co-C bond lengths [1.893 (7) and 1.905 (7) Å ] for compound 1 are as expected for this class of compounds and compare well to values observed by Shores for the cyclam macrocyclic counterpart of 1. (Hoffert et al., 2012) Compound 2 shows characteristics of a trans-influence with elongated Co-C bond lengths [1.927 (2) Å avg.] relative to compound 1. This effect is a result of the stronger -donation from phenylacetylide compared to chloride. The C-C and C C bond lengths of the phenylacetylene ligands fall in the expected region for single and triple bonds respectively. The acetylides in compound 2 show a slightly cumulenic character with elongated C C and shortened C-C bond lengths with respect to compound 1, as was also seen by Shores and coworkers (Hoffert et al., 2012). The Co-N bond lengths for each compound are presented in Tables 1 and 2 and do not deviate significantly from those in previously reported Co tetraazamacrocyclic compounds.

Supramolecular features
The structure of the chlorine salt exhibits monoclinic pseudosymmetry, emulating a double-volume C-centered unit cell with parameters a = 34.721, b = 9.690, c = 15.668 Å , and = 93.41 . The and angles in the monoclinic cell deviate substantially from 90 , being 88.97 and 89.52 when not constrained during data integration. In the crystal structure, the monoclinic pseudo-symmetry manifests itself by the presence of a pseudo b-glide operation along the a-axis of the triclinic cell, Fig. 3. Fig. 4 shows a least-squares overlay of one set of cations A and B, of the surrounding chloride anions and solvate molecules and of a second cation. The pseudo-glide symmetry is mostly obeyed by the constituents of the asymmetric unit; the root-mean-square deviation for one overlaid pair of A  Table 1 Selected geometric parameters (Å , ) for 1.   molecules, for the chloride anions and neighboring cations this is no longer the case. This can especially be seen for a second cation shown in Fig. 4 (on the left), which was not included in the calculation of the least-squares overlay fit, and shows easily recognizable positional shifts for its atoms related by pseudo-symmetry. The substantial deviation of the lattice from the ideal monoclinic symmetry (by 1.03 and 0.48 for and , respectively) leads to an insufficient match and the increased deviations of atoms of the next and second next ions and solvate molecules break the higher symmetry. The crystal under investigation did, however, show signs of slight twinning by pseudo-monoclinic symmetry. The application of a 180 rotation around reciprocal (2 0 1) (command 'TWIN 1 0 0 0 1 0 1 0 1 0 in SHELXL) resulted in a twinning ratio of 0.798 (3):0.202 (3), and R 1 does increase by 2.6% if twinning is ignored during structure refinement.
Overlays of a larger segment of the lattice, along the a and c-axes, are shown in Figs. 5 and 6 (one of the overlaid cells was inverted for this purpose). The overlays are based on a leastsquares fit of the four cobalt ions common to the overlaid structures.
The cations, anions, and solvate molecules in each structure are connected through a series of intermolecular hydrogen bonds (Figs. 7-10, see Tables 3 and 4 for metrical   Displacement ellipsoid plot (set to a 20% probability setting for clarity) for compound 1, showing the pseudo-glide plane perpendicular to the a axis. The shift direction is along b.

Figure 4
Least-squares overlay of one set of cations A and B (to the right) of compound 1. Also shown are the surrounding chloride anions and solvate molecules and a second cation (on the left). Atoms color coded red belong to the original structure and atoms in blue were inverted prior to the least-squares overlay. The least-squares fit is based on all atoms of the cation pair on the right (r.m.s. deviation = 0.138 Å ). For this pair, labels are shown only for the A cation. Labels for atoms of the second pair of cations and for all carbon atoms are omitted for clarity.

Figure 5
Overlays of a larger segment of the lattice of compound 1. One of the overlaid cells was inverted for this purpose. The view is along the a axis of the original cell (blue atoms). The overlay is based on a least-squares fit of the four cobalt ions common to the overlaid structures.

Figure 6
Overlays of a larger segment of the lattice of compound 1. One of the overlaid cells was inverted for this purpose. The view is along the c axis of the original cell (blue atoms). The overlay is based on a least-squares fit of the four cobalt ions common to the overlaid structures. N-HÁ Á ÁN hydrogen bonds with the acetonitrile nitrogen atom, N-HÁ Á ÁO hydrogen bonds to the methanol oxygen, and N-HÁ Á ÁCl hydrogen bonds to both the interstitial chloride anions as well as the cobalt-bound chlorine. The chloride anions are also acceptors for O-HÁ Á ÁCl hydrogen bonds originating from the disordered methanol molecules and for a series of weaker C-HÁ Á ÁCl hydrogen bonds from macrocyclic carbon atoms. The type and number of hydrogen bonds is essentially the same between the two halves of the structure related by pseudo-symmetry, but the exact metrics and numbers are slightly modulated. The N-HÁ Á ÁN, N-HÁ Á ÁO, and N-HÁ Á ÁCl hydrogen bonds, when combined,  Table 3 Hydrogen-bond geometry (Å , ) for 1. Symmetry codes: (i) x; y À 1; z; (ii) Àx þ 1; Ày; Àz þ 1; (iii) Àx þ 1; Ày þ 1; Àz þ 1; (iv) x þ 1; y; z; (v) Àx þ 2; Ày þ 1; Àz; (vi) x À 1; y; z; (vii) Àx þ 1; Ày þ 1; Àz.

Figure 7
Hydrogen-bonding interactions in 1, showing a segment of the ribbons formed by N-HÁ Á ÁN, N-HÁ Á ÁO, and N-HÁ Á ÁCl hydrogen bonds (symbolized by dashed turquois lines). Views are slightly tilted down the a axis (left) and the b axis (right). C-HÁ Á ÁO interactions, omitted for clarity, connect parallel ribbons along the a-axis direction. Disorder of methanol molecules is omitted for clarity.

Figure 8
Hydrogen-bonding interactions in 2, showing both layers connected by N-HÁ Á ÁO, N-HÁ Á ÁF hydrogen bonds. The top layer contains cations and anions of Co1 and S1, respectively, the bottom layer those of Co2 and S2. Hydrogen bonds are shown as dashed turquoise lines. Disorder of one of the trifluoromethanesulfonate anions and methylene chloride is omitted for clarity. View is down the a-axis direction. Table 4 Hydrogen-bond geometry (Å , ) for 2. connect the cations, anions and solvate molecules into ribbons that extend infinitely along the b-axis and are perpendicular to the a-axis, and exactly one unit cell thick in the a-and c-axis directions (Fig. 7). Perpendicular to the a-axis, the ribbons are terminated by methanol O atoms and chloride anions, which at their open sides are surrounded by hydrogen atoms from aliphatic C-H, CH 2 and CH 3 groups, thus connecting parallel ribbons with each other. Perpendicular to the c-axis, ribbons are lined by phenyl and methyl groups from the phenylacetylide and the acetonitrile molecules, respectively. Interactions with neighboring ribbons are van der Waals in nature.
In the triflate salt 2 of the bisacetylide complex, the two cations form N-HÁ Á ÁO and N-HÁ Á ÁF hydrogen bonds with the two triflate anions (Fig. 8). The two molecules have a distinctively different set of hydrogen bonds. The number of hydrogen bonds, their type (N-HÁ Á ÁO versus N-HÁ Á ÁF), and their strength varies substantially between the ion pairs. The first of the two cations, involving nitrogen atoms N1 through N4, features each two N-HÁ Á ÁO and N-HÁ Á ÁF hydrogen bonds (not counting duplicates from triflate disorder), Fig. 9. The second cation, involving nitrogen atoms N5 through N8, makes three N-HÁ Á ÁO hydrogen bonds, and one N-HÁ Á ÁF (Fig. 10). On average, the hydrogen bonds involving this second molecule are much weaker than those involving the first molecule, with two of the N-HÁ Á ÁO bonds and the N-HÁ Á ÁF bond having donor-acceptor distances longer than 3.52 Å . For the first molecule, only one exceeds a value of 3.5 Å , and this one is towards the minor moiety of the disordered triflate anion. The methylene chloride halogen atoms do not act as acceptors for hydrogen bonds, but are involved in weak C-HÁ Á ÁO hydrogen bonds towards one of the triflate anions.

Synthesis and crystallization
All reactions were carried out under ambient conditions. [Co III (DMC)Cl 2 ]Cl was synthesized according to literature procedures (Hay et al., 1984). Hydrogen-bonding interactions in 2, showing the hydrogen-bonded layer formed by cations and anions of Co1 and S1, respectively. Hydrogen bonds are depicted as dashed turquoise lines. View is slightly tilted down the c axis. Disorder of the trifluoromethanesulfonate anion is omitted for clarity.

Figure 10
Hydrogen-bonding interactions in 2, showing the hydrogen-bonded layer formed by cations and anions of Co2 and S2, respectively. Hydrogen bonds are depicted as dashed turquoise lines. View is slightly tilted down the c axis. and Et 3 N (0.77 mL, 5.6 mmol) were added and the solution was refluxed overnight. Solvent was removed via rotary evaporation, and the solid was loaded onto a silica gel plug and eluted with CH 3 OH/EtOAc (v/v, 1:6) as a red fraction. The desired product was recrystallized from ether-methanol to afford 170 mg of a coral solid (73% based on [Co III (DMC)Cl 2 ]Cl). Single crystals were grown from slow diffusion of ether into a methanol solution of 1. Preparation of [Co III (DMC)(C 2 Ph) 2 ]OTf (2). Compound 1 (150 mg, 0.33 mmol) and AgOTf (384 mg, 1.49 mmol) were dissolved in 50 mL of CH 3 CN and refluxed for 48 h. The precipitate that formed was filtered out, and 3.1 mL (22 mmol) of Et 3 N and 0.20 mL (1.8 mmol) of phenylacetylene were added and the solution was refluxed for 48 h. The solution was purified over a silica gel plug and the product eluted with CH 3 OH/EtOAc (v/v, 1:8). A pale-yellow fraction was collected and recrystallized from ether-methanol to afford 102 mg of a yellow solid (47% based on 1). Single crystals were grown from slow diffusion of n-hexanes into a CH 3 OH/ CH 2 Cl 2 (v/v, 1:9) solution of 2.

Refinement
Crystal data, data collection and structure refinement details are summarized in Table 5. H atoms attached to carbon and nitrogen atoms and hydroxyl hydrogen atoms were positioned geometrically and constrained to ride on their parent atoms. Carbon-to-hydrogen bond distances were constrained to 0.95 Å for aromatic C-H. Aliphatic C-H, CH 2 , and CH 3 moieties were constrained to 1.00, 0.99 and 0.98 Å , respectively. N-H distances were constrained to 0.88 Å and O-H distances to 0.84 Å . Methyl and hydroxyl H atoms were allowed to rotate, but not to tip, to best fit the experimental electron density. U iso (H) values were set to a multiple of U eq (C) with 1.5 for OH and CH 3 , and 1.2 for N-H and C-H units, respectively.  The structure of compound 1 exhibits pseudo-symmetry and emulates a double-volume C-centered monoclinic cell in space group C2/c. The pseudo-symmetry is only approximate, and the and angles deviate substantially from the expected 90 for monoclinic (approximate cell dimensions: 34.71, 9.69, 15.67, 88.97, 93.41, 89.52). The structure is, however, twinned by a symmetry operator of the approximate larger monoclinic cell, by a 180 rotation around the [201] direction in reciprocal space (of the actual triclinic cell). Application of the twin matrix 1 0 0, 0 1 0, 1 0 1 yielded a twin ratio of 0.798 (3):0.202 (3).
In the structure of compound 1, each methanol group was refined with two-component disorder with a shared occupancy ratio for the two sites. The C-O bond lengths were restrained to 1.427 (20) Å . Each minor occupancy component was restrained to be similar the respective major occupancy component (SAME command of SHELXL, s.u. = 0.02 Å ). The U ij components for atoms within 2.0 Å were restrained to be similar (SIMU command of SHELXL, s.u. = 0.01 Å 2 ). The alcohol hydrogen atom to neighboring chloride distances were restrained based on hydrogen-bonding considerations. Subject to these conditions, the occupancy rates refined to 0.643 (16) and 0.357 (16).
In the structure of compound 2, the S1 triflate anion was refined with two-component disorder. Each moiety was restrained to have a similar geometry as the S2 triflate anion (SAME command of SHELXL, s.u. = 0.02 Å ). The U ij components for disordered atoms within 2.0 Å were restrained to be similar (SIMU command of SHELXL, s.u. = 0.01 Å 2 ). Subject to these conditions, the occupancy factors refined to 0.503 (22) and 0.497 (22). The dichloromethane molecule was refined with two-component disorder. The minor occupancy component was restrained to have a similar geometry as the major occupancy component (SIMU command of SHELXL, s.u. = 0.01 Å 2 ). The U ij components for atoms within 2.0 Å were restrained to be similar (SIMU command of SHELXL, s.u. = 0.01 Å 2 ). Subject to these conditions, the occupancy factors refined to 0.545 (12) and 0.455 (12).

Chlorido(5,12-dimethyl-1,4,8,11-tetraazacyclotetradecane)(phenylethynyl)cobalt(III) chloride-acetonitrile-
where P = (F o 2 + 2F c 2 )/3 (Δ/σ) max < 0.001 Δρ max = 0.93 e Å −3 Δρ min = −1.40 e Å −3 Special details Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes. Refinement. Refined as a 2-component twin. The structure exhibits pseudo-symmetry and emulates a double the volume C-centered monoclinic cell in space group C2/c. The pseudosymmetry is only approximate, and angles deviate substantially from the expected 90 degrees for monoclinic (approximate cell dimensions: 34.71, 9.69, 15.67, 88.97, 93.41, 89.52). The structure is however twinned by a symmetry operator of the approximate larger monoclinic cell, by a 180 degree rotation around the 2 0 1 direction in reciprocal space (of the actual triclinic cell). Application of the twin matrix 1 0 0, 0 -1 0, 1 0 -1 yielded a twin ratio of 0.798 (3) to 0.202 (3). Each methanol moiety was refined with two component disorder, with a shared occupancy ratio for the two sites. The C-O bond lengths were restrained to 1.427 (20) Angstrom. Alcohol hydrogen atom to neighboring chloride distances were restrained based on hydrogen bonding considerations. Subject to these conditions, the occupancy rates refined to 0.643 (16) and 0.357 (16).

Special details
Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes. Refinement. The S1 triflate anion was refined as two-component disorder. Each moiety was restrained to have to same geometries as the S2 triflate anion. The Uij components for atoms within 2.0 Angstrom were restrained to be similar. Subject to these conditions, the occupancy factors refined to 0.503 (22) and 0.497 (22). The dichloromethane molecule was refined as two-component disorder. The minor moiety was restrained to have the same geometries as the major moiety. The Uij components for atoms within 2.0 Angstrom were restrained to be similar. Subject to these conditions, the occupancy factors refined to 0.545 (12) and 0.455 (12).