Crystal and molecular structures of fac-[Re(Bid)(PPh3)(CO)3] [Bid is tropolone (TropH) and tribromotropolone (TropBr3H)]

The crystal structures of two rhenium(I) complexes were determined and inter- and intramolecular interactions were confirmed with solid-state NMR spectroscopy.

Our focus is to try to understand the basic chemistry (mechanism of action, structure-activity relationships and stability) of these organometallic compounds to aid in the design of new bioactive pharmaceuticals. This also includes the characterization by means of solid-and solution-state multinuclear NMR spectroscopy, single-crystal X-ray diffraction and other spectroscopic methods. The application of solid-state NMR spectroscopy to study hydrogen-bond and other intra-and intermolecular interactions is growing rapidly, with many research groups involved in the development of new techniques to study crystalline and even amorphous phases, making it a useful tool for our purposes as well (Chierotti & Gobetto, 2008;Traer et al., 2007;Zhao et al., 2001;Schutte-Smith et al., 2019a;Wilhelm et al., 2022).

Materials and methods
All reagents employed in the preparation and characterization of the title compounds were of analytical grade, were purchased from Sigma-Aldrich or Merck (South Africa) and were used without any further purification; all experiments were performed aerobically. The IR spectra were recorded at room temperature on a PerkinElmer BX II IR spectrometer in the range 4000-370 cm À1 .
The liquid-state 1 H, 13 C and 31 P NMR spectra were recorded at 25.0 C on a 300 MHz Bruker Fourier NMR spectro-meter, a 400 MHz Avance III NMR spectrometer and a 600 MHz Avance II Bruker spectrometer, respectively, and methanol-d 4 , toluene-d 6 and acetone-d 6 were used as solvents. The chemical shifts () are reported in parts per million (ppm); for methanol-d 4 and acetone-d 6 , the spectra were referenced relative to the solvent peak (3.31 ppm for 1 H and 49.15 ppm for 13 C, and 2.05 for 1 H and 29.92 for 13 C, respectively). Coupling constants (J) are reported in Hz. The solid-state NMR spectra were collected on a 400 MHz Bruker Avance III spectrometer equipped with a 4 mm VTN multinuclear double resonance magic angle spinning probe, operating at 25.0 C. The 13 C NMR spectra were recorded at 100.6 MHz, using the cross polarization magic angle spinning (CP/MAS) technique. A rotating speed of 10000 Hz was used with a contact time of 2 ms, a recycle delay of 5 s and an acquisition time of 33.9 ms. All the spectra were recorded with 3k scans. The samples were packed in 4 mm zirconia rotors.  (Schutte et al., 2012), was dissolved in acetone (30 ml) and triphenylphosphane (32 mg, 0.122 mmol) was added to the solution. The mixture was stirred overnight at room temperature and left to crystallize from the acetone solution (yield: 69 mg, 87%). IR (KBr, cm À1 ): CO = 2010CO = , 1934CO = , 1887 183, 182, 138, 136, 135, 133, 132, 130, 129, 128, 127, 126. 31 (Schutte et al., 2008), was dissolved in acetone (30 ml) and triphenylphosphane (20 mg, 0.0077 mmol) was added to the solution. The mixture was stirred overnight at room temperature and left to crystallize from the acetone solution (yield: 62.5 mg, 91%).

Refinement
Crystal data, data collection and structure refinement details are summarized in Table 1. All aromatic H atoms were placed in geometrically idealized positions (C-H = 0.95 Å ) and constrained to ride on their parent atoms, with U iso (H) = 1.2U eq (C).
In the 1 H NMR spectra, a significant downfield shift is observed from 1 (7.23, 6.91 and 6.84 ppm) to 2 (8.15 ppm) for the tropolonate and tribromotropolonate H atoms, respectively, which is expected due to the electron-withdrawing Br atoms in 2 causing deshielding of the nuclei. This is confirmed in the 31 P NMR spectra with a slight downfield shift in the phosphorus peak of 1 at 18.2 ppm and 2 at 19.6 ppm. The IR carbonyl stretching frequencies of 1 (2010, 1934 and 1887 cm À1 ) are lower than 2 (2018, 1922 and 1889 cm À1 ), which is expected since the tropolonate ligand in 1 is more electron donating than the tribromotropolonate ligand in 2, therefore implying stronger backbonding from the carbonyl ligands to the metal centre and resulting in lower CO stretching frequencies. This, in turn, labilizes the phosphane ligand in the sixth position and is confirmed in the solid-state structures, with the Re-P bond lengths reported as 2.4987 (5) Å for 1 and 2.4799 (11) Å for 2.

X-ray crystallography
A summary of the crystal data for 1 and 2 is given in Table 1. fac-[Re(Trop)(PPh 3 )(CO) 3 ], 1, crystallized in the triclinic space group P1 with one molecule in the asymmetric unit. The molecular diagram and selected bond lengths and angles are given in Fig. 1     and one intramolecular hydrogen-bonding interaction (C-HÁ Á ÁO) are observed in the structure, as well as two intermolecular C-OÁ Á Á and one intramolecularinteraction (Figs. S1 and S2 in the supporting information). A summary of the geometric parameters of these interactions is given in Tables S1 and S2 in the supporting information. Interestingly, the hydrogen-bond interactions involve the tropolonate ligand and the C atoms of the C41-ring as C-H donor atoms, and the O atoms of the tropolonate ring and the O atom of a carbonyl ligand as acceptor atoms. The -interactions, on the other hand, involve interactions between the carbonyl O2 and O3 atoms and the centroids of the five-membered Re/O11/C11/ C12/O12 ring, as well as the arene rings of the phosphane ligand (C21-C26 and C31-C36). fac-[Re(TropBr 3 )(PPh 3 )(CO) 3 ], 2, also crystallized in the triclinic space group P1 with one molecule in the asymmetric unit. The molecular diagram is given in Fig. 1 and selected bond lengths and angles are provided in Table 3. Four intermolecular hydrogen-bond interactions (three C-HÁ Á ÁO and one C-HÁ Á ÁBr) and one intramolecular hydrogen-bond interaction (C-HÁ Á ÁO) are observed in the structure of 2 ( Fig. S3 in the supporting information). A short contact of 3.250 (4) Å is observed between Br2 and O3(Àx + 1, Ày + 1, Àz + 1) (Fig. S4). Two intermolecular contacts form an infinite one-dimensional chain with base vector [110] between Br1 and Br3(x À 1, y À 1, z), and between Br3 and Br1(x + 1, y + 1, z), both with a distance of 3.4809 (7) Å (Fig. S5). A range ofinteractions are observed: one X-HÁ Á Á, twoand three Y-XÁ Á Á interactions ranging between 3.438 (4) and 3.865 (2) Å (Fig. S4). A summary of the geometric parameters of these interactions is given in Tables S3 and S4 in the supporting information. All three Br atoms are involved in short contacts, while Br2 is additionally involved in a -interaction and Br3 is involved as an acceptor in a hydrogen-bond interaction. All five of the ring systems, i.e. the three arene rings of the PPh 3 ligand, the tropolonate ring and the Re1/ O11/C11/C12/O12 five-membered ring, are involved in theinteractions.
The bond lengths and angles of 1 and 2 compare well with each other and also with similar structures in the literature (Gantsho et al., 2020;Schutte-Smith et al., 2019b;Schutte et al., 2007Schutte et al., , 2008Manicum et al., 2020;Bochkova et al., 1987;Kydonaki et al., 2016). The Re-P1 bond length of 1 is slightly longer than in 2, possibly due to the electron-withdrawing effect of the three Br atoms on the backbone of 2. The tropolonate ligand in 1 donates more electron density to the rhenium metal centre, initiating more backbonding from the The Hirshfeld surfaces of 1 and 2, illustrating a curvedness plot (middle), a shape index plot (bottom) and the molecular diagram for clarity of 1 and 2.   Table 3 Selected geometric parameters (Å , ) for 2.
carbonyl ligands, labilizing the Re-P bond. Although this is what we expect, it is not observed in the Re-CO bond lengths of 1 and 2, which do not differ significantly. Considering the angles around the Re I metal core, a good correlation between 1 and 2 is found. The small bite angles of 73.99 (5) and 73.05 (10) for 1 and 2, respectively, indicate the degree of distortion of the octahedral geometry, which is normal and within the range of other similar structures where a fivemembered O,O 0 -chelate ring is formed with the metal centre (Gantsho et al., 2020;Schutte et al., 2007Schutte et al., , 2008Schutte-Smith et al., 2019b;Bochkova et al., 1987). In the case of a sixmembered O,O 0 -chelate ring (with PPh 3 in the sixth position), the bite angle is slightly larger, with values ranging between 82.2 and 84.7 (Manicum et al., 2020;Kydonaki et al., 2016). The tropolonate and tribromotropolonate ligands bend slightly towards the triphenylphosphane ligand in 1 and 2, with dihedral angles between the plane through the Re(CO) 3 entity and the ligand (the plane through Re/C1/O1/C2/O2 and the plane through O11/O12/C11-C17) of 8.85 (8) and 12.43 (14) , respectively (illustrated in Fig. S6 in the supporting information). In 2, the Br atoms are slightly 'out of plane' with respect to the tropolonate ring (C11-C17) at À0.1463 (4), 0.1760 (5) and À0.2114 (5) Å for Br1, Br2 and Br3, respectively. This could be due to the different interactions observed: the intermolecular contacts between Br1 and Br3 and the C-HÁ Á ÁBr3 hydrogen-bond interaction, and the Br2Á Á ÁO3 short contact and C15-Br2Á Á ÁCg1(Àx + 1, Ày + 1, Àz + 1) -interaction for Br2 (Cg1 is the centroid of the Re/C1/O1/C2/ O2 ring).
Overall, the curvedness of 1 has less 'flat' regions compared to 2, and compares well with the increased number of -interactions observed in 2 compared to 1.
When d norm (as defined and explained by Spackman & Jayatilaka, 2009) is mapped on a Hirshfeld surface, intermolecular contacts appear as red spots, contacts shorter than van der Waal separations, on a largely blue surface. It has been proven to be useful as an unbiased method to identify close intermolecular contacts, even in complex crystal structures. d norm Hirshfeld plots of 1 and 2 are presented in Fig. 5, indicating the red spots associated with close contacts. Not all interactions are shown for conciseness because all the interactions are not visible from one orientation. All the interactions reported in Tables S1-S4 correlate with these plots.
By comparing 1 and the previously reported bis(triphenylphosphane) complex [Re(Trop)(PPh 3 ) 2 (CO) 2 ] (3) (Gantsho et al., 2020), it is clear that most of the bond lengths around the metal centre change when the axial carbonyl ligand is substituted by a second PPh 3 ligand (Table 4). When the carbonyl ligand is substituted by a PPh 3 ligand, more electron density is donated to the Re I metal centre, shortening the equatorial Re-CO bond lengths from 1.900 (2) and 1.912 (2) Å to 1.883 (3) and 1.887 (3) Å . PPh 3 also has a weaker trans effect than CO, which is evident in the shortening of the Re1-P1 bond(s).
Interestingly, the trans effect is clearly observed in the axial Re-CO distances in the solid-state crystal structures of fac- The atom-numbering schemes of 1 and 2.

Solid-state NMR
In solid-state 13 C NMR spectroscopy, the cross polarization magic angle spinning (CP/MAS) technique is often used to enhance the polarization of the low-abundance 13 C nuclei via its interaction with 1 H nuclei. The effectiveness of the CP/ MAS technique depends on the magnitude of 1 H-13 C dipolar coupling (Freitas et al., 2016;Conte et al., 2004;Smernik et al., 2002). It is expected that the observed hydrogen-bond interactions, as well as other short contacts and -interactions in the solid state, will deshield the C atoms and cause a downfield shift in the solid-state 13 C NMR spectra (Patterson-Elenbaum et al., 2006). In the liquid state, the intra-and intermolecular interactions are disrupted because of the motion of the molecules within the solution; thus, we only observe the dynamic average of the motion. The degree of interactions present in the solid-state can be determined by the difference in chemical shift values (Á) of the specific C atoms in the liquid-versus solid-state NMR spectra (Patterson-Elenbaum et al., 2006). A larger difference in chemical shift is normally indicative of a stronger interaction, which is determined by the specific bond length and angle (Siskos et al., 2017).
It is known that broad peaks (or no peaks) are observed when there are not many C atoms that are directly bound to H atoms (Freitas et al., 2016), which is the case in 2. Nevertheless, we aimed to correlate the change in chemical shift from the 13 C liquid-state NMR to the solid-state 13 C NMR to the interactions observed in the crystal structures. Fig. 6 provides the numbering scheme of atoms in 1 and 2. The solid-state 13 C NMR data of 1 did not shift much from the solution state to the solid state, with not more than a 1 ppm change (Á) in the chemical shift at most, which is basically negligible (Fig. 7). Four hydrogen-bond interactions and three -interactions are observed in 1 (Tables S1 and S2 in the supporting information), two of the -interactions being very weak (distance > 3.8 Å ). The five interactions that are considered to be stronger with shorter distances involve the PPh 3 ligand, the O atoms of the tropolonate ligand, the carbonyl ligands and the centroid of the five-membered Re1/ O11/C11/C12/O12 ring system. The carbonyl ligands are not visible on the liquid-and solid-state 13 C NMR spectra due to the economic and time implications involved to observe it. IR spectroscopy are used to confirm the presence of the carbonyl ligands in this type of complex.
In the solution-state 13 C NMR spectra, the peak at 184 ppm is assigned to C11 and C12, and seeing that these atoms are bound to O11 and O12 (involved in three interactions) and are part of the five-membered ring system (Re1/O11/C11/C12/ O12), a downfield shift is expected because of the effect of the Solution-state versus solid-state 13 C NMR spectra of 2.
deshielding of these interactions. However, it shifted slightly upfield to 183 and 182 ppm, and yielded two peaks in the solid state compared to a single peak in the solution state. This is due to the fact that C11 and C12 are not equivalent in the solid state because of the interactions observed in the crystal structure: C17-H17Á Á ÁO1(Àx + 2, Ày + 1, Àz), C45-H45Á Á ÁO11(Àx + 2, Ày + 1, Àz) and C46-H46Á Á ÁO11 all indirectly involve C11, while C44-H44Á Á ÁO12(x, y + 1, z) is the only interaction that indirectly involves C12; thus, the splitting of the single peak in the solid state.
The single peak for C13, C14, C16 and C17 at 138 ppm, the peaks for the PPh 3 ligand at 134, 131 and 129 ppm, and the single peak for C15 at 127 ppm in the solution state are not as well defined in the solid state and yield a broad peak from 138 to 126 ppm, similar to the range found in the solution state; we expected a downfield shift because of the interactions involving the tropolonate ring system and one arene ring of the PPh 3 ligand. We could, however, see some significant splitting of the peaks; compared to the five single peaks at 138, 134, 131, 129 and 127 ppm in the solution state, splitting of the peaks (although it is a broad peak) is seen in the solid state, indicating that many of the C atoms are not equivalent anymore because of the interactions observed in the crystal structure [C17-H17Á Á ÁO1(Àx + 2, Ày + 1, Àz), meaning C13, C14, C16 and C17 are not equivalent anymore; C44-H44Á Á ÁO12(x, y + 1, z), C45-H45Á Á ÁO11(Àx + 2, Ày + 1, Àz) and C46-H46Á Á ÁO11, meaning the C41-C46 arene ring in PPh 3 is not equivalent to the C21-C26 and C31-C36 arene rings].
In the case of 2, the fact that the tribromotropolonate ring system only has two H atoms directly bound to C atoms had an impact on the solid-state 13 C NMR spectra and we only observe the PPh 3 ligand, and the seven C atoms in the tribromotropolonate ligand are not observed (Fig. 8) (Freitas et al., 2016). In the solution-state spectra, the PPh 3 ligand has a single peak at 134 ppm which split up into four peaks at 135, 132, 131 and 129 ppm in the solid-state spectra. Again, this is because the C atoms are not equivalent in the solid state.

Conclusion
Two new crystal structures of rhenium(I) tricarbonyl complexes with either a tropolonate or a tribromotropolonate bidentate ligand are reported and correspond well with similar known structures. The solid-state NMR data indicated the presence of inter-and intramolecular interactions, as seen by the splitting of some signals, but unfortunately, due to the fact that both 1 and 2 contain only a few C-H units each, credible chemical shifts could not be obtained and correlated with the crystal data. The intermolecular interactions obtained from PLATON (Spek, 2020)

fac-Tricarbonyl(triphenylphosphane-κP)(tropolonato-κ 2 O,O′)rhenium(I) (1)
Crystal data 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 reflection data of fac-[Re(Trop)(PPh 3 )(CO) 3 ] and fac-[Re(TropBr 3 )(PPh 3 )(CO) 3 ] were collected at 100 (2) K on a Bruker D8 Quest Eco Chi Photon II CPAD diffractometer using Mo Kα radiation (λ = 0.71073 Å) and at 104 (2) K on a Bruker D8 Venture 4K Kappa Photon III C28 diffractometer also using Mo Kα radiation (λ = 0.71073 Å), respectively. The unit-cell parameters were refined by SAINT-Plus (Bruker, 2012), while SADABS (Bruker, 2012) was used for absorption corrections. The structures were solved by direct methods and refined on F 2 using anisotropic displacement parameters for all non-H atoms. SHELXL97 (Sheldrick, 1997(Sheldrick, , 2008 and WinGX (Farrugia, 2012) were used for structure solutions and refinements, respectively. The molecular graphics were prepared with DIAMOND (Brandenburg & Putz, 2019).

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å 2 )
x y z U iso */U eq Re1 0.90667 (2) 0.18057 (2)    where P = (F o 2 + 2F c 2 )/3 (Δ/σ) max = 0.001 Δρ max = 0.93 e Å −3 Δρ min = −1.48 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.  (2)